Citation
Inheritance of resistance to the soybean looper in soybean

Material Information

Title:
Inheritance of resistance to the soybean looper in soybean
Creator:
Kenty, Michael Montgomery, 1959-
Publication Date:
Language:
English
Physical Description:
x, 102 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Breeding ( jstor )
Crops ( jstor )
Defoliation ( jstor )
Estimation methods ( jstor )
Insects ( jstor )
Leaf area ( jstor )
Memory interference ( jstor )
Modeling ( jstor )
Soybeans ( jstor )
Statistical discrepancies ( jstor )
Agronomy thesis Ph.D
Dissertations, Academic -- Agronomy -- UF
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 95-101).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Michael Montgomery Kenty.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030599382 ( ALEPH )
32021356 ( OCLC )
AKG3548 ( NOTIS )

Downloads

This item has the following downloads:


Full Text









INHERITANCE OF RESISTANCE TO THE SOYBEAN LOOPER
IN SOYBEAN















By

MICHAEL MONTGOMERY KENTY


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1994






























I would like to dedicate this work to the memory of the late Dr. Johnny D. Ouzts.

Thank you for your friendship and guidance; without it I would not have made it this far.














ACKNOWLEDGEMENTS


My deepest gratitude is extended to Dr. Kuell Hinson for serving as my mentor

throughout my program of study. I would also like to thank the remaining members of

my graduate committee, Drs. Joe Funderburk, Ken Quesenberry, John Strayer, and

David Wofford, for their insight and guidance. This study could not have been

completed without the support and aid of the USDA-ARS Soybean Production Research

Unit at Stoneville, Mississippi. I want to thank each and every member for their

friendship, patience, and support during this educational process. Also, I would like to

thank Dr. Clarence Watson and Debbie Boykin for their statistical advice. A special

thanks goes to Drs. Edgar E. Hartwig and Thomas C. Kilen for taking a chance on a

young kid years ago, and introducing me to the wonderful field of plant breeding.














TABLE OF CONTENTS


page


ACKNOWLEDGEMENTS .................

LIST OF TABLES ......................

LIST OF FIGURES ......................

ABSTRACT .............. ..........

CHAPTER 1 INTRODUCTION .............

CHAPTER 2 LITERATURE REVIEW .........


. . iii

. vi

. viii

. ix

S. 1


CHAPTER 3 EVALUATION OF RATING METHODS .............. 13
Introduction ......................... ......... ............ 13
Literature Review ..................... ............. 13
Materials and Methods ....... ......................... 15
Results and Discussion .................................. 19
Summary and Conclusions ........... ................... 29

CHAPTER 4 INHERITANCE OF RESISTANCE TO SOYBEAN LOOPER IN
SOYBEAN ............................. ............. 31
Introduction .................................. ............ 31
Materials and Methods ............... ................. 32
1988 . . 32
1989 . . 33
1990 . .. 34
1991 ... . .. 39
Results and Discussion ................................. 49
1990 .. . 49
1991 . . 51
Summary and Conclusions ....... ............. ......... 66

CHAPTER 5 SUMMARY AND CONCLUSIONS .................... 71
Evaluation of Rating Methods ....... ........... ......... 72
Inheritance Study ............... ..... ................. 74








APPENDIX A LIST OF DEFOLIATION SCORES GENERATED BY
WHOLE PLANT VISUAL, PARTITIONED PLANT VISUAL, AND
LEAF AREA OF THE PARTITIONED PLANT RATING METHODS
IN 1990. .......................................

APPENDIX B LIST OF DEFOLIATION SCORES GENERATED BY
WHOLE PLANT VISUAL, PARTITIONED PLANT VISUAL, AND
LEAF AREA OF THE PARTITIONED PLANT RATING METHODS
IN 1991. ................................. ........


APPENDIX C


GENERATION MEANS ANALYSIS SAS PROGRAM .....


APPENDIX D GENERATION MEANS ANALYSIS SAS PROGRAM .....


APPENDIX E


TEST FOR NORMALITY ......................


REFERENCES ......................................... 95

BIOGRAPHICAL SKETCH ................................. 102













LIST OF TABLES


page

Table 1. McNemar test for significant differences between the WP (whole plant
visual) and AV (partitioned plant visual) methods in 1990; and between
the WP and AV, WP and AL (leaf area of the partitioned plant), and AV
and AL methods in 1991 ............................... 28

Table 2. Populations used for generation means analysis of the inheritance of
resistance to the soybean looper. .... ....... 46

Table 3. Ratings of leaf feeding by soybean looper on soybean parents, their F,
and F2 progeny, and the germplasm lines PI229358 and D75-10169 in the
field cage-1990..................................... 50

Table 4. Goodness of fit test for normality of soybean looper defoliation for
populations of D86-3429, Braxton, and D86-3429 x Braxton F2
populations in the 1990 cage study. ......................... 52

Table 5. Ratings of leaf feeding by soybean looper on soybean parents, their F1,
F2, F3, BCI, BSI, BC2, and BS2 progeny, and the germplasm lines PI
229358 and D75-10169 in the field cages-1991. ............... 54

Table 6. Estimates of variance components obtained from an analysis of the
homogeneous populations, D86-3429, Braxton, D75-10169, PI 229358,
for a location effect due to feedingt by soybean looper in field cages in
1991. .......................................... 55

Table 7. Means, standard errors, and variances of soybean looper defoliation
ratings for parental lines and seven populations of soybean arising from
the cross D86-3429 x Braxton grown in field cages at Stoneville, MS in
1991. . . 57

Table 8. Mather's scaling test applied to the cross D86-3429 x Braxton to test
adequacy of additive-dominance model for resistance to soybean looper. 58








Table 9. Estimates of the additive, dominant, and epistatic effects in the
generation means for defoliation by soybean looper in the nine populations
of the D86-3429 x Braxton material grown in the field cages at
Stoneville, MS in 1991. ................... ............ 60

Table 10. Estimates of the additive, dominant, and epistatic effects in the
generation means for defoliation by soybean looper in the nine populations
of the D86-3429 x Braxton material grown in the field cages at
Stoneville, MS in 1991. ................... ............ 61

Table 11. Goodness of fit test for normality of soybean looper defoliation data
from 1991 cage study ................................. 63

Table 12. Average defoliation scores in soybean populations resistant to soybean
looper in 1991 cage study. ................... ........... 67














LIST OF FIGURES


page

Figure 1. Examples of photocopied leaflets used in determining the % defoliation
in the AV (leaf area of partitioned plant) method ..... 17

Figure 2. 1990 frequency distribution of defoliation estimates by the whole plant
and partitioned plant average visual rating methods. ... 21

Figure 3. 1991 frequency distribution of defoliation estimates by the whole plant,
partitioned plant average visual, and measured average leaf area of the
partitioned plant rating methods. ... .. ...... 22

Figure 4. Schematic drawing of field cages used in the 1990 and 1991
experiment ... ..... 35

Figure 5. Mating scheme to derive populations required to estimate the genetic
effects of resistance to the soybean looper in soybean. ... 45













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of
Doctor of Philosophy

INHERITANCE OF RESISTANCE TO THE SOYBEAN LOOPER
IN SOYBEAN

By

Michael Montgomery Kenty

April, 1994

Chairman: Kuell Hinson
Major Department: Agronomy

Phytophagous insects cause millions of dollars of damage to soybean [Glycine max

(L.) Merr.] throughout the southern United States. This study was conducted to: (i)

evaluate three methods for rating defoliation of soybean by insects and (ii) determine the

inheritance of resistance to the soybean looper [Pseudoplusia includes Walker] in

soybean. The insect-resistant breeding line D86-3429 was crossed with the cultivar

Braxton in 1988. Additional crosses were made in 1989 and 1990 to provide the

populations necessary to conduct rating methodology and genetic studies in field cages

in 1990 and 1991. The three rating methods evaluated were whole plant visual (WP),

partitioned plant average visual (AV), and measured average leaf area of partitioned plant

(AL). In 1990 the WP and AL were highly correlated (r=0.65, p < 0.001) and in 1991

all correlations (WP vs. AV, WP vs. AL, and AV vs. AL) were significant at p <








0.001. Based on the relative variation of each method, the precision of the three

methods was essentially equal and very good, well below the desirable level of 10.

These results indicated that relative estimates are suitable for genetic studies. In the

inheritance study defoliation was estimated by the WP method in 1990 and the AV

method in 1991. The data from the preliminary study in 1990 showed a trend towards

quantitative inheritance, therefore the 1991 data were analyzed quantitatively. Mather's

scaling test was applied to the data generated from each cage and the results indicated

that generation means depend only on additive gene effects. Utilizing Hayman's

methodology in terms of the generation means analysis, an epistatic effect was suggested,

but the primary effect is assumed to be additive, as was indicated in the scaling test.

Estimates of gene numbers indicate that the two parents differed by two genes for

resistance to soybean looper. Heritability for resistance was estimated to be 63%,

therefore a breeder should be able to make progress by selecting in the F2 or F3

populations. The potential exists for increasing resistance to the soybean looper, and

possibly other phytophagous insects, by pyramiding genes from new sources of

resistance.













CHAPTER 1
INTRODUCTION



Soybean [Glycine max (L.) Merrill] is a major world crop. Total world hectares

planted increased from 47.1 million in 1977-78 to 57.7 million in 1989-90 (2,3). This

22.5% increase was due primarily to demand for oil and meal products, and to a small

extent the whole bean product (75). In the United States, hectarage rose to a peak of

28.6 million hectares in 1978-79 with a steady decline to 24.0 million hectares in 1989-

90. This 16% decrease in soybean hectarage was primarily due to reduced plantings in

the southeastern United States (3).

The reduction in hectares planted to soybean in the southeastern United States is

attributable to the low market value of soybean. Another reason for the reduction is the

inability of farmers, in general, to effectively produce a profitable crop in the presence

of yield-reducing factors. These factors include fungal diseases, viral diseases, bacterial

diseases, nematodes, insects, and drought. In 1982, Mississippi suffered an estimated

$50 million loss in revenue from insects alone, of which $40 million could be attributed

solely to the soybean looper, Pseudoplusia includes Walker (15). Similar losses also

occurred in states neighboring Mississippi as a result of insect infestations (15,16). Pest

problems which can be costly to control, and a 10-year (1981-1990) average price of

$6.19 per bu (3), make it increasingly difficult for a grower to produce a profitable crop.








2
Several techniques are available for control of diseases, nematodes, and insects

in soybean. Metcalf and Luckmann (50) categorized these techniques into seven major

methods: cultural, mechanical, physical, biological, chemical, genetic, and regulatory.

All methods of control have merit and should be utilized in integrated pest management

programs (14,20,58,50). Although economics is the major limiting factor in achieving

control of any pest of soybean, human safety and environmental impact must be

considered when choosing an approach to control a particular pest. Before making a

decision on what approach to take, a grower should answer the questions: 'Will the end

results justify the cost of control?' and 'Is the control measure safe?'.

The soybean looper has become increasingly resistant to available insecticides

(10,18,48,58,60). For control of the soybean looper and other insect pests, a biological

approach should yield the best results. There are two biological approaches to insect

control: host plant resistance and the use of biocontrol agents (14,58). The latter

involves the identification and exploitation of natural enemies and pathogens. This

method of control usually is limited to a specific insect pest rather than the broad

spectrum of insects that damage soybean. As with conventional insecticides, resistance

to microbial insecticides has been observed (58).

Host plant resistance offers an effective method of control which is built into the

plant. It has relatively long stability, is compatible with other control tactics, is

environmentally safe (58), and development and subsequent release of a resistant cultivar

is far more socially acceptable (82). Lambert and Kilen (47) found that direct selection

for resistance in soybean to one species of insect resulted in the indirect selection for








3
resistance to certain other species. The development of an insect-resistant soybean

cultivar takes approximately the same amount of time as is required to develop a new,

effective insecticide, which can be as short as 10 to 12 years or as long as 20 years. The

insect-resistant cultivars 'Lamar' (28) and 'Crockett' (5) each took approximately 20

years to develop.

Initially, Van Duyn et al. (83) identified the three plant introductions PI 171451,

PI 227687, and PI 229358 from the USDA germplasm collection as having the highest

level of resistance to the Mexican bean beetle (Epilachna varivestis Mulsant). Schillinger

(69) identified additional sources of resistance to specific insects and described the

mechanism of resistance that each exhibits. Hatchett et al. (30) determined that breeding

lines identified as possessing resistance to multiple species of insects could serve as basic

germplasm for development of cultivars resistant to insects. To date, two soybean

cultivars and two soybean germplasm lines with resistance to insects have been registered

with the Crop Science Society of America (5,11,28,29).

Although there has been success in developing insect-resistant germplasm lines

and cultivars, there has been very little research conducted on the inheritance of insect

resistance in soybean (37,54,72). Of the three inheritance studies, two (54,72) were

conducted on the resistance to Mexican bean beetle and one (37) on the resistance to the

velvetbean caterpillar (Anticarsia gemmatalis Hubner). Since the soybean looper is a

major pest to soybean in the southeastern United States and is difficult to control with

insecticides, the inheritance of resistance to the soybean looper would aid breeders in the

development of resistant cultivars.








4
The objectives of this study were: (i) to evaluate three methods for rating

defoliation of soybean by soybean looper and (ii) to determine the inheritance of

resistance in soybean to the soybean looper.













CHAPTER 2
LITERATURE REVIEW



Although undocumented, plants that are resistant to insects have survived and

propagated based on the processes of adaptation and natural selection. These 'survivors'

were in turn selected by farmers to plant future crops (74). This practice of selection by

primitive farmers could be considered the beginning of the development of insect-

resistant cultivars.

According to Kogan (38), Pedigo (58), and Smith (74), in their brief histories of

insect-resistant plants, the first development and subsequent cultivation of insect-resistant

cultivars occurred in the late eighteenth and early nineteenth centuries. In 1792, J.N.

Havens identified the wheat (Triticum aestivium L. em. Thell.) cultivar 'Underhill' as

being resistant to the Hessian fly (Mayetiola destructor Say). In the 1830's, G. Lindley

recommended the cultivation of apple (Malus pumila Miller) cultivars resistant to the

woolly apple aphid (Eriosoma lanigerum Hausman). In the mid-nineteenth century, the

French wine industry was saved by grafting the highly susceptible scions of the French

varieties to the rootstocks of American grapes (Vitis L. spp.), which were resistant to the

grape phylloxera (Phylloxera vittifolae Fitch).

In the United States, R.H. Painter is considered to be the founder and pioneer of

the modern era of research on insect-resistant plants (58,74). In his book, Insect








6
Resistance in Crop Plants, which was the first book written on the topic, he described

mechanisms of resistance in plants and factors that affect them. Based on field

observations, he separated the mechanisms of resistance in plants into three distinct, but

interactive classes: nonpreference, later known as antixenosis (38), (for oviposition, food,

or shelter), antibiosis (adverse effect on the biology of the insect), and tolerance (ability

to withstand, repair, or recover from infestation) (57). Developing a cultivar that

exhibits a high level of all three mechanisms of resistance would be optimum. This is

seldom feasible, therefore developing cultivars with the antibiosis mechanism of

resistance is the most frequent objective of plant breeders. Although nonpreference and

antibiosis differ in their mechanisms of resistance, both involve the interaction of the

biology of the plant with the biology of the insect (58,74). There is often an overlapping

of the nonpreference and antibiosis classes which makes it difficult to determine the exact

mechanism of resistance exhibited in a plant to a particular species of insect (74).

Although it is advantageous if a breeder knows which mechanism of resistance

is being expressed, knowledge of the mechanism is not essential to make advances in the

development of resistant germplasm. In fact, resistant cultivars and germplasm have

been developed for years with little concern for the underlying mechanism of resistance.

It has only been in the past two decades that mechanisms of resistance have been

identified in crop plants (39).

In soybean, Van Duyn et al. (83) identified three genotypes (PI 171451, PI

227687, and PI 229358) from the USDA germplasm collection of maturity groups VII

and VIII as resistant to the Mexican bean beetle. These three PI's have been used








7
extensively as sources of resistant germplasm and as standard checks in screening

additional germplasm for resistance. At the World Soybean Research Conference in

1975, Schillinger (69) reported additional sources of resistance to the Mexican bean

beetle. Gary et al. (22) screened 1108 lines from maturity groups VI, VII, VIII, and IX

of the USDA germplasm collection against velvetbean caterpillar initially, and

subsequently velvetbean caterpillar, soybean looper, corn earworm (Heliothis zea

Boddie), tobacco budworm (H. virescens Fabricius), and beet armyworm (Spodoptera

exigua Hubner). Based on the results of the screenings against these five species, PI

209837 and FC 31592 were identified as having levels of resistance that would be

suitable for developing resistant cultivars. The germplasm collection of 473 PI's of wild

soybean, Glycine soja Sieb. & Zucc., was screened against soybean looper, velvetbean

caterpillar, beet armyworm, and corn earworm to identify additional sources of resistance

equal to, or greater than, PI's 229358 and 171451. McKenna et al. (52) identified three

lines (PI's 464935-1, 366119, and 407301) from this study as having less defoliation than

the two standards, thus suggesting that there are additional genes for insect resistance.

In 1988, Kraemer et al. (42) identified several accessions of the USDA germplasm

collection as being resistant to the Mexican bean beetle. They were from maturity

groups VI (FC 31665, and PI's 379621, 416925, and 416937) and VIII (PI's 417061 and

417136).

Upon identification of a genotype that is resistant to a certain insect species, the

genotype usually is screened against other insect species. Resistance to more than one

species will enhance its usefulness in a breeding program. Hatchett et al. (30) conducted








8
a study utilizing breeding lines that derive their insect resistance from PI 229358 to

determine if selection for resistance to one or two insect species would result in the

indirect selection for resistance to other troublesome species. They found that no single

insect species could be utilized to select for resistance to multiple species of insects

among adapted genotypes. In a similar study, Lambert and Kilen (47) determined that

when PI 229358 was used as the donor parent, direct selection for resistance to one

insect species resulted in indirect selection for resistance to other foliar-feeding species.

Research to develop soybean genotypes resistant to insects is usually conducted

with species that are indigenous to a particular area (19,25,35,44,54,64,80,83). Lambert

and Kilen (45) evaluated PI229358, PI227687, and PI 171451 against five insect species

to determine their relative levels of resistance. Insect species tested were velvetbean

caterpillar, soybean looper, corn earworm, tobacco budworm, and beet armyworm.

Based upon the responses to these five species, they found that the mechanism for

resistance in PI 171451 was nonpreference, whereas that of PI 229358 or PI 227687 was

antibiosis. These same three PI's were evaluated against four important soybean

defoliating species (beet armyworm, Porthesia taiwania Shiraki, Anomala cupripes Bates,

and Orgyia species) to determine their usefulness in developing breeding lines resistant

to insects indigenous to Taiwan. No single accessions had a consistently high level of

resistance to all of the insect species. PI 227687 demonstrated the highest level of

resistance to the most prevalent species, the beet armyworm, and therefore would be

suitable to initiate a breeding program (80).








9
With the sources of resistance to various species of insects identified, researchers

have channelled their energy into determining the basis of the mechanism of resistance

in certain soybean accessions. Since the mid-1970's, researchers have investigated

various constituents of soybean to identify the underlying biochemical compound that

confers resistance to insects (9,53,73,81). If an association could be established between

a particular compound or class of compounds and resistance, then a new tool could be

developed to accelerate the selection process by breeders.

In studies with PI 227687, PI 229358, and two susceptible cultivars, Tester (81)

reported that resistant PI's had lower total nitrogen, more soluble carbohydrates, and

more total sterols at flowering and pod-fill than susceptible cultivars. Through a series

of grafting experiments, Lambert and Kilen (46) verified that constituents contributing

to resistance in PI 227687 and PI 229358 were formed in the leaves and translocated

throughout the plant. Smith (73) determined that the resistant factors in PI 227687

appeared to be chemical. He also found that resistance to the soybean looper in PI

227687 was an antibiotic effect due to the combined effects of a feeding deterrent and

a growth inhibitor. Chiang et al. (9) suggested that resistance to Mexican bean beetle

feeding in PI 227687 may be attributed to phenylpropanoid metabolites. In their

inorganic nutrient analysis of leaf tissue studies with soybean resistant to the Mexican

bean beetle, Mebrahtu et al. (53) found that the content of calcium, phosphorus, and

potassium was negatively correlated with pupal weight gain (r = -0.23", r = -0.26",

and r = -0.20'; p < 0.01 and p < 0.05, respectively).










In 1986, Kogan (39) published a treatise on the role of biochemicals in plant

resistance to insects. He examined the defensive patterns of three families: Solanaceae,

Cruciferae, and Leguminosae (specifically the common bean and soybean) and published,

as quoted below, a summation of all previous research findings that elucidated the

chemical bases of resistance in PI 171451, PI 227687, and PI 229358.

1. Foliar total nitrogen, soluble carbohydrates, organic acids
and sterols of both resistant and susceptible soybean vary
with stage of growth (Tester, 1977; Grunwald and Kogan,
1981).

2. Total nitrogen was generally lower and, at the end of the
season, soluble carbohydrates and sterols were higher in
the resistant lines; organic acids fluctuated without
correlation with resistance (Testor, 1977).

3. Sterol profiles were similar in both susceptible and resistant
lines, both in early and in late maturing soybeans
(Grunwald and Kogan, 1981).

4. Pinitol is a dominant cyclitol in soybean and it represents
an average of 60% of the total 80%-ethanol-soluble
carbohydrate fraction of soybean (Phillips et al., 1982).

5. Pinitol was found in large amounts in PI 229358, and in
smaller concentrations in the susceptible cultivar 'Davis'
(@ 1% dry weight). Added to artificial media at 0.8%,
pinitol reduced weight gain by corn earworm larvae by
about 63%. These findings suggest that pinitol is an
antibiotic factor in soybean but with no apparent antixenotic
property (Dreyer et al., 1979). As the reported results
could not be replicated elsewhere, this claim has been
recently disputed (Gardner et al., 1984).

6. Systematic fractionation of PI's 227687 (Smith and Fischer,
1983) and 229358 (Binder and Waiss, 1984) yielded
fractions with intermediate and with polar solvents that
caused slower developmental rates and high mortality in
soybean looper [Pseudoplusia includes (Walker)] and
Heliothis zea Boddie, respectively.









7. Seven phenolic acids were detected in PI 171451 and in cv.
'Forrest'; gallic, protocatechuic, p-hydroxybenzoic,
vanillic, caffeic, p-coumaric, and ferulic. The potentially
antibiotic caffeic and ferulic acids occurred at higher
concentrations in the resistant PI than in the susceptible
cultivar. In both types of plants, concentrations were
higher in injured than in uninjured tissue (Hardin, 1979).

8. Isoflavonoid phytoalexins in soybean cotyledons are potent
feeding deterrents for the Mexican bean beetle (Hart et al.,
1983). These phytoalexins are biochemically related to the
phenolics mentioned above, through the shikimic pathway.
(39, p.516-518).

Although a great deal of knowledge has been gained, there still remain many unanswered

questions as to what components comprise the various mechanisms of resistance.

A moderate amount of research has been conducted in the area of resistance in

soybean to insects, and much progress has been made in cultivar development.

However, there has been very little information published on the inheritance of resistance

to insects in soybean. In their work with the Mexican bean beetle, Sisson et al. (72)

hypothesized that resistance was inherited quantitatively and that the number of major

genes involved was small, perhaps two or three major genes. In further studies with the

Mexican bean beetle, Mebrahtu et al. (54) concluded that resistance appeared to be

primarily an expression of additive genetic variance. Kilen et al. (36), in their studies

with the soybean looper, suggested that susceptibility was partially dominant with a few

major genes conditioning resistance. In their studies with PI 171451, PI 227687, and PI

229358, and the F, plants derived from intercrosses among the three PI's, Lambert and

Kilen (45) suggested that genetic resistance in PI 171451 differed from that operating in

the other two lines. This was further supported by their work with the PI's and the F3








12
lines derived from the intercrosses among the PI's. They hypothesized that each PI

contained a resistance gene that differed from those of the other lines. To design a better

breeding program to develop cultivars resistant to the soybean looper utilizing breeding

lines that derive their insect resistance from PI 229358, it would be advantageous if the

mode of inheritance of this trait were known.














CHAPTER 3
EVALUATION OF RATING METHODS


Introduction

Throughout the past 25 years there have been advances in the identification and

development of soybean breeding lines resistant to phytophagous insects, especially

Lepidopterous species and the Mexican bean beetle (5,11,22,28,29,30,42,52,79,83).

Despite these advances, little effort has been devoted to improving the methods of

evaluating the amount of feeding injury of insects. Most researchers utilize a rating

system that estimates the level of resistance among cultivars, but little or no effort has

been made to determine whether or not the rating scales used were truly reflecting the

defoliation caused by insects. Hence, the objective of this study was to evaluate the

effectiveness of the widely used visual, or relative, assessment of insect defoliation as

compared to a quantitative, or absolute, assessment.

Literature Review

Plant pathologists have faced problems in the assessment of disease which is

dependent upon the causal organism and the host plant species. Many different methods

of disease assessment are available to plant pathologists (7). Shokes et al. (71) tested the

reliability of disease assessment procedures with late leafspot [Cercosporidium

personatum (Berk. and Curt.) Deighton] in peanut (Arachis hypogaea L.) and concluded








14

that plant pathologists would benefit if a standard rating system for foliar diseases of

peanut could be adopted as has been done in other crops. In a special report, Berger (4)

stated that a disease assessment procedure should have the following basic requirements:

easy to learn, quick and efficient to use, applicable across a broad spectrum of

conditions, and be accurate, precise, and reproducible. He concluded that the greater the

emphasis placed on the assessment method, the higher the quality of information obtained

by the method.

In 1980, Kogan and Turnipseed (41) described the various methods of quantitative

and qualitative assessments of defoliation that are currently used by researchers working

with phytophagous insects. These methods are: (i) photoelectric or electronic (measure

actual leaf area), (ii) gravimetric (regressing known leaf areas on the corresponding fresh

or dry weight of these areas, (iii) volumetric (volume displacement cylinder is used to

measure biomass), (iv) planimetric (mechanical measurement of area), (v) geometric

(estimates total nondefoliated leaf area so that percent defoliation can be calculated when

used in conjunction with a leaf area meter), and (vi) visual estimates (provide rapid

estimate of defoliation). Although descriptions of the methods were detailed by Kogan

and Turnipseed, no statistical comparisons were made among methods.

Some authors (58,78) refer to qualitative assessments as relative estimates and

quantitative assessments as absolute estimates. Relative estimates do not translate directly

into the amount of defoliation, whereas absolute estimates allow for direct translation into

the amount of defoliation caused by phytophagous insects. The major drawback of

absolute estimates is that they are usually time consuming and costly to make.








15
Conversely, relative estimates can be timely and are inexpensive which is why these

types of estimates are more widely used by researchers (74).

Funderburk et al. (19) determined that plot size or shape had little effect on the

relative rankings of soybean cultivars when measuring resistance to the velvetbean

caterpillar in terms of percentage of defoliation based on visual estimates. However,

they found that the mean and precision of the estimates were affected by plot size and

shape. The objective of this study was to compare two relative methods to one absolute

method for assessing defoliation caused by the soybean looper.

Materials and Methods

The three rating methods evaluated in 1990 and 1991 were the whole plant visual

rating (WP), the partitioned plant average visual rating (AV), and measured average leaf

area of the partitioned plant (AL). WP values were based on a 1-to-10 scale divided into

10% increments with a score of 1 signifying defoliation of 0 to 10% and a score of 10

signifying defoliation of 91 to 100%. Although other scales, such as a logarithmic scale

(33), may be better at estimating defoliation, the 1-to-10 scale was chosen since it is most

commonly used by participants in the Regional Host Plant Resistance Test which is

sponsored annually by the Southern Regional Information Exchange Group (SRIEG 32).

With the AV and AL methods, the canopies were visually partitioned into three sections

(top, middle, bottom) enabling a more detailed assessment of soybean looper defoliation

on each plant. The partitioning of the plants was based on a procedure that was

developed by Plaut and Berger (59) that was cited by Shokes et al. (71). For AV








16
estimates, each section of the plant was scored using the same scale as the WP with the

three scores averaged to obtain the AV.

In 1990, a LI-COR LI-3000 (Li-Cor, Inc., Lincoln, NE 68504) portable leaf area

meter was used to measure the middle leaflet of a representative trifoliate from each of

the sections and the three measurements were averaged to obtain the AL. In 1990 the

AL method only allowed an indirect comparison to the other two methods, therefore the

AL was modified in 1991 to permit a direct comparison with AV and WP. In the 1991

AL method, a representative trifoliolate leaf from each section of the partitioned plant

was removed and taken into a laboratory to determine the area of each sample. The

portable leaf area meter was used to determine the remaining portion of each defoliated

leaflet (DEF) of each trifoliolate. Following the procedure described by Kogan and

Turnipseed (41), each leaflet of the trifoliolates was then placed on a photocopy machine

and a copy was made. Examples of leaflet photocopies are shown in Figure 1. These

silhouettes were then cut out to represent the trifoliolates in a nondefoliated state and the

area (NONDEF) for each was determined as mentioned above. The two values obtained

for each trifoliolate were used in the following formula to determine the percent

defoliation (% def): (NONDEF DEF)/NONDEF 100 = % def. The percent

defoliation for each of the three trifoliolates from each plant was then converted to the

same 1-to-10 scale used by the other two methods and the values averaged, allowing

direct comparisons to be made among the three methods.

The plants used in the evaluation of the three methods were part of an inheritance

study that was conducted in field cages (43) in 1990 and 1991 at the Mississippi State




















































Figure 1. Examples of photocopied leaflets used in determining the % defoliation
in the AV (leaf area of partitioned plant) method.








18

University Delta Research and Extension Center, Stoneville, Mississippi. In the first

year of the study, seed were planted on 25 May 1990. In 1991, seed were planted on

25 May and 29 May in cage 1, and on 31 May in cage 2. Rainfall on 25 May prevented

completion of planting on the same day in cage 1.

To compare methods in 1990, ten soybean plants were chosen randomly from

each of 20 backcross families for a total of 200 plants. Fifty plants in each of two cages

were randomly chosen using PROC PLAN of PC-SAS (68) for use in the evaluation of

the three methods in 1991. These plants were chosen prior to release of soybean looper

moths in both years.

The 4000 three-day-old soybean looper moths were released on 25 June 1990

when plants were at the V5 to V6 stage of plant growth (17). Larvae from hatched eggs

were allowed to feed freely for approximately 12 days, at which time 'Centennial' (27),

the susceptible check, was estimated to have a rating of nine according to the WP

method. The area within the cage was then sprayed with methomyl [S-methyl-N-((methyl

carbamoyl) oxy)-thioacetimidate] at 0.5 kg ha' to kill the larvae. This allowed

defoliation ratings to be made without further feeding. The day following spraying, each

plant was rated by all three methods.

Approximately 4000 three-day-old soybean looper moths were released into each

cage on 8 July 1991. At the time of release, plants in cage 1 were V7 to V8 and in cage

2 they were V6 to V7. The larvae from the hatched eggs were allowed to feed freely

until the susceptible check, Centennial, received a rating of ten by the WP method, 16

days after release of the moths. To cease feeding at this stage of plant growth, R1 to R2








19

in cage 1 and beginning R1 in cage 2, the plants in both cages were sprayed with

methomyl at 0.5 kg ha"' to kill the larvae. The following day each of the 100 plants was

rated using the WP and AV methods and samples were collected for the AL.

Data derived from each rating method for the 200 plants in 1990 and the 100

plants in 1991 were subjected to analytical procedures in PC-SAS (68). A correlation

analysis was run to determine the correlation coefficient (r) between WP and AL, AV

and WP, and AL and WP rating methods each year. A paired difference comparison

(76) was made between WP and AV methods in 1990 and between WP and AL, WP and

AV, and AL and AV methods in 1991. McNemar's test for concordance (77) was used

to test for significant changes between WP and AL in 1990 and between WP and AL,

WP and AV, and AL and AV methods in 1991. This analysis was conducted to

determine if the methods differed in their estimates of defoliation. Precision, known as

the relative variation (RV), of the two relative methods for both years and the absolute

method in 1991 were calculated for each as follows:


RV = (SE/.)100


where SE is the standard error of the mean and X is the mean. This relationship,

described by Ruesink (66) and Pedigo (58), is considered useful in estimating precision

of a sampling method by entomologists.

Results and Discussion

The 1990 scores for each of the three rating methods by plant are presented in

Appendix A. Scores for AV and WP were based on a 1-to-10 scale, with lower numbers








20
indicating less defoliation. The measurements for AL in 1990 are presented as cm2 of

remaining tissue. Individual plant data for each of the three methods in 1991 are

presented in Appendix B. Unlike 1990, 1991 data for all three methods were based on

a 1-to-10 scale. Means and standard deviations for both years are also reported in

Appendices A and B.

Frequency distributions of the 1990 WP and AV rating methods are shown in

Figure 2. The AL method was omitted from this graph because it measured the

remaining undefoliated tissue in cm2, whereas the AV and WP methods were used to

estimate percent defoliation on a 1-to-10 scale. Frequency distributions of the AV and

WP methods suggest that either the evaluator underestimated defoliation by the WP

method or overestimated defoliation by the AV method. Without an absolute estimate

of defoliation, no substantial conclusion can be made as to the reliability of AV or WP

to accurately assess defoliation.

The 1991 frequency distributions of all three methods for both cages combined

are presented in Figure 3. Due to the modifications in the AL method, it was possible

to make direct comparisons between AL and the visual rating methods. The evaluator

overestimated defoliation to obtain too few plants below 30% or <4 in the 1-to-10 scale

when using either relative method, WP or AV, as compared to AL, the absolute method.

Also, the evaluator had a tendency to overestimate the amount of defoliation over 50%

or >5 in the 1-to-10 scale when using either relative method as compared to AL. Only

in classes 4, 5, and 6 was the evaluator's estimate of defoliation with the WP and AV

methods similar to that of the AL. The problem of overestimating might be avoided in









21




Number of Plants
20
Rating Methods
*WP HAV

60 -




40




20




1 2 3 4 5 6 7 8 9 10
Defoliation Scores



Figure 2. 1990 frequency distribution of defoliation estimates by the whole plant and
partitioned plant average visual rating methods. Defoliation scores based on 1-to-10
scale where 1 = 0 to 10% defoliation, 2 = 11 to 20%, 3 = 21 to 30%, 4 = 31 to 40%,
5 = 41 to 50%, 6 = 51 to 60%, 7 = 61 to 70%, 8 = 71 to 80%, 9 = 81 to 90%, 10
= 91 to 100%.














Number of Plants


Rating Methods
MWP HAV MAL


1 2 3 4 5 6 7 8 9 10


Defoliation Scores


Figure 3. 1991 frequency distribution of defoliation estimates by the whole plant,
partitioned plant average visual, and measured average leaf area of the partitioned plant
rating methods. Defoliation scores based on 1-to-10 scale where 1 = 0 to 10%
defoliation, 2 = 11 to 20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to 50%, 6 = 51
to 60%, 7 = 61 to 70%, 8 = 71 to 80%, 9 = 81 to 90%, 10 = 91 to 100%.








23
the future if a logarithmic rating scale, such as the Horsefall-Barratt scale to assess

disease (33), were used to estimate defoliation. A logarithmic scale may have advantages

for assessment of defoliation because the visual acuity of the human eye is more adapted

to the identification of small differences at each end of the scale.

The plant scores of the AV and the WP rating methods in 1990 were highly

correlated (r = 0.65, p < 0.001). AL measurements were in cm2 of nondefoliated leaf

area remaining, therefore not a measure of defoliation. As there was no absolute

measure of defoliation, it was not possible to make further comparisons with AV and

WP.

In 1991, the scores for cage 1 from the AV and WP methods were significantly

correlated with the AL method (r = 0.86, p < 0.001; and r = 0.88, p < 0.001,

respectively). The AV and WP methods correlated well with each other (r = 0.88, p

< 0.001). By squaring each of the correlation coefficients (r), both the WP vs. AV and

WP vs. AL relationships had r2 values of 0.77 meaning 77% of the variance in one rating

method can be explained by the variation in the other. The AV vs. AL correlation had

a r2 value of 0.74, indicating that 74% of the variance of the AV method can be

explained by the variation in the AL method (or vice versa).

Although the results from ratings on plants in cage 2 were similar, correlations

were not as high as those in cage 1. Scores from the AV and WP methods were

significantly correlated with the AL method (r = 0.75, p < 0.001; and r = 0.67, p <

0.001, respectively). The AV and WP methods were significantly correlated (r = 0.74,

p < 0.001) with each other. Calculated r2 values were: WP vs. AV r2 = 0.55, WP vs.








24
AL r2 = 0.49, and AV vs. AL r2 = 0.56. These values are lower than those found with

the cage 1 data. Differences may be attributed to differences in the plant growth at the

time of rating or to evaluator error. The cage 2 ratings were made in the rain which

hindered the evaluator's ability to accurately estimate defoliation. It was considered

more prudent to make ratings in the rain than to attempt to disregard the ensuing

regrowth which was increasing rapidly during a prolonged rainy period.

When data for both cages were pooled for analysis, scores for the AV and WP

methods were significantly correlated with the AL method (r = 0.79, p < 0.001; and

r = 0.78, p < 0.001, respectively). The AV and WP methods correlated well with each

other (r = 0.80, p < 0.001) as in the cages treated separately. The following r2 values

were obtained: WP vs. AV r2 = 0.64, WP vs. AL r = 0.60, and AV vs. AL r2 = 0.62

thus approximately 62% ( 2%) of the variation in one rating method could be attributed

to the variation in the other rating method.

In a similar study, a relative method and an absolute method were compared using

their estimates of defoliation of field plots infested primarily (90% of total insect

population) with the velvetbean caterpillar (G.R. Bowers, 1992, personal

communication). A 1-to-9 scale, with 1 signifying the least damage, was used as the

relative method. The absolute method followed a procedure described by Kogan and

Kuhlman (40) in which 20 leaflets each from the top, middle, and bottom thirds of plants

(for a total of 60) were sampled from plants located in the center two rows of a four-row

plot. These were then compared to diagrams of defoliated leaves to determine the

defoliation estimate of each sample. The two methods were significantly correlated, r








25
= 0.62, p < 0.001. Bower's correlation coefficients are somewhat similar to the pooled

correlation coefficients between the relative and absolute methods in this 1991 cage

experiment.

Correlations from the 1991 experiment, whether examined individually by cage

or combined, were all significant at p < 0.001. The magnitude of the r2 values indicate

that the evaluator obtained similar relative rankings of defoliation regardless the method

used. Although the correlations between the AL method and the WP and AV methods

were highly significant, it was not possible to determine if the evaluator estimated the

same amount of defoliation with the WP and AV methods (relative methods) as with the

AL method (the absolute method). To compare the accuracy of the evaluator using the

two relative methods and the absolute method, additional analyses were undertaken.

With the PROC MEANS procedure of PC-SAS, a paired difference comparison

was made between the WP and AV methods in 1990. A t-test was performed to test the

equality of the two methods based on the differences between the two methods for each

plant in the study. The mean difference of -0.36 between the WP and AV methods was

highly significant, p < 0.001, meaning the WP method yielded lower average percent

defoliation than the AV method. However, without some measure of the actual or

absolute defoliation across the plants, there was no way to determine if the evaluator

could obtain an accurate assessment of the true defoliation of a plant with either method.

A paired difference comparison was also made among the three rating methods

in 1991. The WP and AV methods had a mean difference of -0.55, p < 0.001, which

was somewhat similar in magnitude to the mean difference calculated in the 1990 study.








26
The mean differences standard deviation involving the AL method were 1.48 0.14

for WP AL and 2.03 0.12 for AV AL, respectively. Both differences were

significant at the p < 0.001 level. Although the evaluator estimated a higher average

percent defoliation with both relative methods than with the absolute method, the estimate

obtained with the WP method was numerically more similar to the measured defoliation

obtained with the AL method.

With the relative methods, WP and AV, the evaluator estimated somewhat similar

amounts of defoliation across all plants when compared with each other, even though the

difference between the two methods was highly significant. However, in comparisons

with an absolute method, AL, it becomes apparent that the evaluator's estimates of

defoliation with the WP method were closer to the actual than were the estimates of the

AV method. Although the AL method is accepted as the absolute measure of defoliation

for each plant, there is some evidence that it underestimated defoliation. Necrotic leaf

tissue due to insect injury was present. The leaf area meter could not distinguish viable

leaf tissue from necrotic tissue. The amount of necrotic tissue was not enough to

increase AL values to the level of WP value, however the best absolute estimate of

defoliation would have been to sample the entire plant (destructive sampling) and follow

the protocol of the AL method after cutting out necrotic tissue. In the 1991 experiment,

provisions were not made for destructive sampling; therefore it was not practiced.

McNemar's test for concordance was used to determine if the categorization of

plants having less than or equal to 50% defoliation and of plants having greater than 50%

defoliation differed with the WP and AV methods. In 1990, 36 of the 198 of the plants








27
were classified with the WP method as having at most 50% defoliation while with the

AV method these same plants were rated as having greater than 50% defoliation (Table

1). This percentage (18%) of plants was significantly different from zero, X2 = 34.03,

p < 0.001, supporting the earlier contention (Figure 2) that either the WP method

underestimated defoliation or the AV method overestimated defoliation. Without the

absolute method for reference it was not possible to determine which method was more

accurate.

All three methods were compared in 1991 to determine whether significant

differences among the methods existed. The results from McNemar's test for

concordance of the 1991 data are shown in Table 1. The two relative methods were

significantly different (X2 = 11.25, p < 0.001) with 20 of the 97 plants being classified

differently by WP than AV. The relative methods were both significantly different from

the absolute method (WP vs. AL X2 = 22.04, p < 0.001; and AV vs. AL X2 = 38.03,

p < 0.001) in the estimate of defoliation on a plant-by-plant basis. With both relative

methods, the evaluator had a tendency to overestimate defoliation as compared to the

absolute method. With the AV method 40 of 97 plants were classified as > 50%

defoliated while AL classified these same plants as < 50%, whereas with the WP

method only 24 of 97 plants were classified > 50% as compared to the AL method

classifying them as 5 50%. This indicates that the evaluator was able to estimate

defoliation closer to the absolute method with the WP method as was seen in the paired

difference comparison.










Table 1. McNemar test for significant differences between the WP (whole plant visual) and
AV (partitioned plant visual) methods in 1990; and between the WP and AV, WP and AL
(leaf area of the partitioned plant), and AV and AL methods in 1991.

1990
WP
AV 1 5t > 5t Totals X2
1-5 159 0 159
>5 36 3 39
Totals 195 3 198 34.03"'
1991
WP
AV 1-5 > 5
1-5 39 2 41
> 5 18 38 56
Totals 57 40 97 11.25'"
WP
AL 1-5 > 5
1-5 57 24 81
> 5 0 16 16
Totals 57 40 97 22.04'"
AV
AL 1-5 > 5
1-5 41 40 81
> 5 0 16 16
Totals 41 56 97 38.03**
*** Significant at the 0.001 level of probability.
t Defoliation ratings based on a 1-to-10 scale where 1 = 0 to 10% defoliation, 2 = 11 to
20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to 50%, 6 = 51 to 60%, 7 = 61 to 70%, 8
= 71 to 80%, 9 = 81 to 90%, 10 = 91 to 100%.








29
RV values for 1990 were 2.11 and 2.77 for WP and AV, respectively. According

to Pedigo (58), RV values less than 10 are desirable in research, but the closer the value

is to zero the more precise the method is at making estimates. Although the WP method

was numerically better than the AV method, the precision of both methods was

essentially equal and very good.

The RV values for each of the methods in 1991 were as follows: WP = 4.26, AV

= 3.49, and AL = 4.09. Based on the RV values, the AV method was numerically

better than either WP or AL. The AL method was expected to be the most precise since

it was the absolute measure of defoliation. However in this study, these values were

essentially equal and close to zero, thus the evaluator was fairly precise in making

estimates of defoliation with all three methods.



Summary and Conclusions

The objective of this study was to evaluate the effectiveness of the widely used

visual, or relative, assessment of insect defoliation as compared to a quantitative, or

absolute, assessment. In 1990 no absolute method comparisons were possible, however

the WP and AV methods were highly correlated (r = 0.65, p < 0.001). The evaluator

estimated a slightly lower amount of defoliation on the average with the WP method than

with the AV. The relative variation of these two relative methods were both less than

3.0 which according to Pedigo (58) is very desirable. Without an adequate absolute

method no determination could be made as to which relative method best reflected the

actual defoliation.








30
Adjustments in the method of calculating percent defoliation by the AL method

for 1991 allowed for direct evaluations of the two relative methods. All correlations

(WP vs. AV, WP vs. AL, and AV vs. AL) were significant at the p < 0.001 level,

whether the cage data were treated separately or pooled. The two relative estimates of

defoliation were significantly correlated with the absolute estimate of defoliation. These

results are similar to the findings of a another relative versus an absolute estimation of

defoliation conducted in 1985 by G.R. Bowers (G.R. Bowers, 1992, personal

communication).

Paired comparisons were made among the methods to test the equality of the

methods to rate defoliation. In 1991, the estimates of defoliation by all three methods

were significantly different. These findings were validated by McNemar's test for

concordance. The WP defoliation estimates were more similar to the AL results, than

the AV ratings were to AL.

Based on the RV's of each method, the precision of the AV and WP methods was

essentially equal in 1990. Similar results were obtained in 1991 with the AV, WP, and

AL methods. In both years the precision of the methods was essentially equal and well

below the value of 10 that Pedigo (58) states as being desirable. These results indicate

that either relative method is suitable for estimating defoliation.

The results of these studies demonstrate that relative estimates of defoliation can

be used effectively in genetic studies or routine screenings of advanced breeding lines.

Since the two relative methods yielded similar results, it would be advantageous to use

the most economical method.












CHAPTER 4
INHERITANCE OF RESISTANCE
TO SOYBEAN LOOPER IN SOYBEAN


Introduction

During the past three decades great strides have been made in the identification

and development of insect-resistant breeding lines of soybean (22,30,42,52,79,83). This

research has led to the release of several germplasm lines and cultivars (5,11,28,29) that

are resistant to phytophagous insects. Although advances have been made, very little

information has been published on the inheritance of resistance to foliar-feeding insects.

Sisson et al. (72) reported that resistance to the Mexican bean beetle was

quantitatively inherited with two or three major genes involved. In other inheritance

studies, Mebrahtu et al. (54) concluded that resistance to the Mexican bean beetle was

primarily an expression of additive genetic variance. Kilen et al. (36) suggested that

susceptibility to the soybean looper was partially dominant and a few major genes

conditioned resistance. Based on their work with PI 171451, PI 227687, and PI229358

and the FI progeny derived from the intercrosses among the three PI's, Lambert and

Kilen (45) suggested that the genetic basis for insect resistance in PI 171451 differed

from that in the other two PI's. This was further supported by their work with the three

PI's and the F3 lines derived from intercrosses among the three PI's (37). They

hypothesized that each PI contained a distinct resistance gene.








32
To facilitate the design of breeding programs to develop cultivars resistant to the

soybean looper, it would be beneficial to understand more completely the mode of

inheritance of this trait. Therefore, the objective of this study was to determine the mode

of inheritance of resistance to the soybean looper in genotypes derived from crosses

involving PI 229358.

Materials and Methods

All aspects of this experiment were conducted at the Mississippi State

University, Delta Research and Extension Center at Stoneville, MS on a Bosket fine

sandy loam (Mollic Hapludalfs).

1988

To study the inheritance of resistance to soybean looper in soybean, the breeding

line D86-3429 was crossed with the cultivar Braxton (32). D86-3429 has white flowers

(wl), gray pubescence (t), sensitivity to metribuzin [4-amino-6-(1,1-dimethylethyl)-3-

(methylthio)-1,2,4-triazin-5,(4H)-one] (hm), resistance to Phytophthora rot (induced by

Phytophthora megasperma Drechs. f. sp. glycinea T. Kuan & D.C. Erwin) (Rpsl-c), and

resistance to soybean looper. The donor of insect resistance was the germplasm line

D75-10169 (29), which derived its resistance from PI 229358 (83). The male parent,

Braxton, has purple flowers (WI), tawny pubescence (7), tolerance to metribuzin (Hm),

susceptibility to Phytophthora rot (rps), and susceptibility to soybean looper. Both

parents were of maturity group VII. Hand pollinations (1,60) made in the field produced

seven seeds of the cross D86-3429 x Braxton.










1989

The seven F, seeds produced in 1988 were space-planted 30.5 cm apart in 91.5-

cm rows in the F, nursery. Several rows of each parent, D86-3429 and Braxton, were

planted at a rate of 14 seeds per meter for crossing purposes. Four F, plants emerged

and these plants were used as the pollen parents for two backcrosses, D86-3429 x

F,(D86-3429 x Braxton) (BC1) and Braxton x FI(D86-3429 x Braxton) (BC2). The

original cross, D86-3429 x Braxton, and its reciprocal were also made to provide

additional F, seed for the following year. Approximately 60 hand pollinations were made

for the two backcrosses and the two crosses with the expectation that an adequate number

of seed would be produced. The two backcrosses, BC, and BC2, produced fourteen and

six seed, respectively. At maturity, the seed from the four F, plants were harvested to

provide F2 seed for use in 1990. In addition, seed from both parents were harvested for

use in 1990. Seven seeds were produced from the cross D86-3429 x Braxton and five

seeds were produced from the reciprocal cross. The seed of the backcrosses, BC, and

BC2, were scarified, treated with metalaxyl [N-(2,6-dimethylphenyl)-N-

(methoxyacetyl)alanine methyl ester] to optimize rapid emergence of healthy plants, and

planted one seed per pot in 15 L pots filled with soil treated with Bradyrhizobium

japonicwn inoculum in the greenhouse during the winter of 1989-90. Plants were

allowed to self-pollinate, thus providing seed of the BS, (the selfed generation of BC,)

and BS2 (the selfed generation of BC2) generations. At maturity the plants were

harvested and threshed with an ALMACO (Allen Machine Co., Nevada, IA 50201)

single-plant thresher.









1990

The inheritance study was conducted in a screened field cage 2.4 m high x 19.2

m wide x 32.0 m long (43). A schematic of the field cage design is shown in Figure

4. In mid-April the soil was disked twice in opposite directions with a disk harrow. On

23 May the soil was tilled with a spring-tooth harrow to smooth the surface. This was

followed with a preplant-incorporated herbicide treatment of clomazone 2-(2-

chlorophenyl) methyl-4,4-dimethyl-3-isooxazolidinone and trifluralin [2,6-dinitro-N,N-

dipropyl-4-(trifluoromethyl)benzenamine] at 1.12 kg ha-' and 0.84 kg ha', respectively,

on 24 May. The area within the cage was marked with rows spaced 68.5 cm apart to

be used as a guide for planting with push planters.

The seed for the study were planted on 25 May. A susceptible cultivar,

'Centennial' (27), was planted at 27 seed per meter of row in border rows as well as a

1 m section at the ends of each row of experimental material. Seed to produce

experimental plants were scarified, treated with metalaxyl, and space planted 25 cm

apart. The parents, D86-3429 and Braxton, were planted in three 12.7 m rows for a

total of 150 plants each. Also, 50 seed from each of the fourteen BC, and six BC2 plants

were planted in 12.7 m rows. Following the backcross progeny, five F, seed each of the

cross D86-3429 x Braxton and its reciprocal were planted. These were followed by the

360 F2 seed from the F, (D86-3429 x Braxton) plants grown in 1989. The insect-

resistant germplasm lines D75-10169 (29) and PI 229358 (83) were planted in the middle

and at the end of the experimental material for a total of 70 plants each.































/


iLJ'


i


,It


r
i
I
i








36
Normal production practices were used to maintain healthy plants. Soil was

cultivated on 7 and 20 June to control weeds, and bentazon [3-(1-methylethyl)-(1H)-

2, 1,3-benzothiadiazin-4(3H)-one 2,2-dioxide] was applied at 0.84 kg ha' on 19 June with

a hand held pneumatic sprayer. Supplemental irrigation was not required during this

time. A Saran cage top (43) was installed on 21 June as the final step in preparation for

the release of insects.

Approximately 4000 laboratory-reared (26) soybean looper moths were released

in the cage at 0830 h on 25 June. The absence of physical restrictions made all areas of

the cage equally accessible to the moths. Water was applied by furrow irrigation to

elevate the relative humidity within the cage to enhance egg laying. All plants were at

the V5 to V6 growth stage at the time of release. Larvae emerged from eggs after an

incubation period of approximately three days.

Larvae were allowed to feed until susceptible Centennial was determined to be 85

to 90 % defoliated which occurred on 6 July or 11 days after moths were released. At

that time, methomyl was applied at 0.5 kg ha' with a backpack sprayer to kill the

loopers and stop further feeding. On 9 July, the Saran cage top was removed to facilitate

regrowth by plants that were to be harvested for use in 1991, and all plants in the study

were evaluated for defoliation. Each plant was assigned a score of 1-to-10 (AV method)

based on the visual estimate of defoliation. The 1-to-10 rating scale represented 10%

increments, with a score of 1 representing defoliation of 0 to 10% and a score of 10

representing defoliation of 91 to 100%.








37

Ratings for the two parents, D86-3429 and Braxton, and their F, and F2 progeny

were grouped by their defoliation scores and plotted on a graph. The distribution of the

parental and F, populations were tested for normality following the procedure described

by Leonard et al. (49). Initially the defoliation data of these populations were

summarized in frequency distributions by using the upper class limits. The mean i,

variance s2, and standard deviation s of each population were calculated from the class

centers (i.e. 0.5, 1.5, 2.5, ... 9.5) using the following formulas:


S= n1(0.5) + n2(1.5) + ... + nio(9.5)/(ni + n2 + ... + n0)

s2 = n(0.5 i) + n1(1.5 i) + ... + nj(9.5 i)/(n.. 1)

s = (s2)12


where n, was the number of plants in each of the respective classes and n.. was the grand

total. Following the calculation of these variables, the variable x was calculated for each

class in the following manner:


x = (i upper class limit)/s


After determining the x for each class, a table of normal probability integrals was used

to determine the corresponding Z value of each class. The theoretical percentage of each

population was calculated by subtracting Z from 1.000 and multiplying by 100. The

theoretical number of each class was calculated as follows:


n.. x theoretical % = theoretical n








38
For each succeeding class, the cumulative percentage in all prior classes must be

subtracted to obtain the theoretical percentage to calculate the theoretical n of that class.

The observed distribution versus theoretical distribution was tested with a X2 test. This

was done to gain insight into the mode of inheritance of resistance to soybean looper and

to facilitate the design of the 1991 experiments. The information from these four

populations was used to determine what additional crosses, if any, would be required to

fulfill the objectives of this study.

The original cross, D86-3429 x Braxton, was repeated to generate seed to be

used as the FI generation in 1991. The two backcrosses, BCi and BC2, were repeated

for the same purpose as the original cross, as well as to provide seed for planting in a

winter nursery to generate the BSi and BS2 populations, the selfed generations of the

backcrosses. The bulk of the pollinations were made on plants within the cage. The

remainder were on plants located in the F2 nursery. Plants within the cage were irrigated

as needed throughout the remainder of the growing season to insure adequate seed set on

F1 and F2 plants for use the following year.

Twelve seeds were produced from the D86-3429 x Braxton cross. The two

backcrosses, BC, and BC2, yielded 49 and 29 seeds, respectively. The 12 F, seeds were

combined with remnant D86-3429 x Braxton F, seed generated from the 1988 and 1989

crosses for use in 1991. Twelve seeds and 11 seeds, from BC, and BC2, respectively,

were scarified and treated with metalaxyl and sent to the winter nursery at Mayaguez,

Puerto Rico to produce the BSi and BS, generations. The single plants from the various

generations in the cage were harvested and threshed with a single-plant thresher between








39
31 October and 2 November. The soil in the cage was subsoiled on 9 November in

preparation for the 1991 experiment.

1991

Two field cages were used to complete the inheritance study. Soil in each cage

was disked with a disk harrow and treated with trifluralin at 0.84 kg ha-, which was

incorporated with a spring-tooth harrow on 26 April. To control a broad spectrum of

weeds, a tank mix of clomazone at 1.12 kg ha' and imazaquin [2-(4,5-dihydro-4-methyl-

4-(1-methylethyl)-5-oxo-1H-imidazol-2-yl)-3-quinolinecarboxylicacid] at 0.11 kg ha1 was

applied and incorporated with a spring-tooth harrow on 13 May. On 25 May, soil was

reworked with a spring-tooth harrow and rows spaced 68.6 cm apart were marked to

serve as guides for the push planter.

Each row consisted of 10 plants spaced 25 cm apart except for the Fi, BC,, and

BC, generations, which had only one plant per row. There were three rows, consisting

of 10 plants each, of the insect-resistant germplasm lines D75-10169 and PI 229358.

The planting order for each cage was generated by randomizing the 277 rows with PC-

SAS PROC PLAN (68). Each cage contained the entire complement of plants and was

treated as a replication over environments.

The border rows were planted to Centennial at 27 seed per meter in each cage.

The rows of experimental populations were centered 68.6 cm apart. Rainfall of 43 mm

halted planting on 25 May in cage 1 after one third of the rows were planted. The

remainder of cage 1 and all of cage 2 was lightly tilled with a spring-tooth harrow on 29

May. The remaining rows in cage 1 were planted on 29 May. Approximately 1 m of








40
Centennial was planted at the ends of all rows to serve as a border. The experimental

populations were surrounded by a susceptible cultivar.

On 31 May, soil in cage 2 was rolled with a water-filled roller pulled by a four-

wheel all-terrain vehicle. This operation smoothed and firmed the seed bed, which

decreased the rate of evaporation of moisture from the soil. The rows were then marked

as in cage 1. The susceptible cultivar Centennial was planted by the same method as in

cage 1. The entire area within cage 2 was planted on 31 May. Short sections of the

Centennial border rows were replanted on 14 June.

Soil in the cages was cultivated on 4, 14, 20, and 26 June with a self-propelled

garden tiller. Supplemental soil moisture was provided by furrow irrigation on 4 and 21

June to ensure that the plants would have sufficient vegetative growth at the time of

insect release. The Saran tops were put on the cages on 26 June as final preparation for

the release of insects.

Approximately 4000 laboratory-reared (26) soybean looper moths were released

in each cage at 0800 h on 8 July. A rainfall of 19 mm on 4 July provided sufficient soil

moisture to maintain an elevated relative humidity within the cages during the egg-laying

period. At the time of release, the plants were at the V7 to V8 stage of growth in cage

1, and V6 to V7 in cage 2 (17). Larvae emerged after an incubation period of

approximately three days.

Larvae were allowed to feed in each cage until 22 July, at which time Centennial

was approximately 95% defoliated. The cage tops were then removed to facilitate the

defoliation ratings and to promote regrowth of the plants. Both cages were sprayed with








41

methomyl at 0.5 kg ha'- to kill larvae. This insured that additional feeding would not

occur once the ratings had begun. Cage 1 was rated on 23 and 24 July using the AV

rating method. This method used the same 1-to-10 scale previously described for the WP

method in 1990. In the AV method the plants were visually partitioned into three

sections (top, middle, bottom). Each section was scored for defoliation and these three

scores were averaged to get the mean score for each plant. This method enabled the

evaluator to make a more detailed assessment of defoliation by the soybean looper on

each plant. Cage 2 was rated during a rainfall by the same method on 25 and 26 July.

All ratings in both cages were made by the same evaluator.

Generation means and variances were calculated for each of the nine generations

on a cage basis. These means and variances were employed in Mather's scaling tests

(51) to determine the adequacy of an additive dominant model and to test for epistasis.

The formulas for the different scaling tests were as follows:


A = 2RBC1 XPI RxF VA = 4ViBCI + ViPI + ViFI

B = 2xBC2 RP2 F, VB = 4VxBC2 + VXP2 + ViFi

C = 4XF2 2iF, ixP iP2 Vc = 16ViF2 + 4ViFi + ViP, + ViP,

D = 8RF3 3iPi 3iP2 2iF1 VD = 64VRF3 + 9VixP + 9VPP2 + 4VRFI


where i was the mean of the respective generations and Vi was the variance of the mean

of the same respective generations. The variance of the mean of each generation was

calculated as Vi = o2/n, where o2 was the variance of the generation and n was the

number of individuals observed in that generation.








42
To determine if each of the scaling tests had the expected value of zero, each was

tested in the following manner:

tA = A/(VA)C

t, = B/(VB,)

tc = C/(Vc)'

to = D/(VD)H

where (VA)", (VB)', (Vc)', and (VD)" were the standard errors of each tests. The t

values obtained were compared to a table of t distribution (76) to determine the level of

significance for each of the tests. The probability was found in the table of t using the

sum of the degrees of freedom of each generation in each test as the number of degrees

of freedom.

Since the preliminary results indicated that resistance to soybean looper is

inherited quantitatively, it was necessary to select an appropriate method of analysis that

would allow for a detailed study of the inheritance of resistance. After reviewing the

statistical methods available for the analysis of a quantitative trait, a generation means

analysis was selected. This analysis was chosen over other methods because of the

following advantages (23):

1. Errors are inherently smaller since means (first order

statistics) are used, rather than variances (second order

statistics);

2. Smaller experiments allow the same degree of precision;










3. Additive (a), dominant (d), and epistatic (aa, ad, and dd)

effects are estimated using means rather than variances;

and

4. It is applicable to both cross- and self-pollinating crops.

In utilizing the generation means analysis, it is assumed that the parents are homozygous

and there is no linkage among genes influencing the trait being studied. If this

assumption is correct and it appears that the trait in question is not quantitatively

inherited, these generations can be used to analyze for Mendelian ratios.

Hallauer and Miranda (23) cited the following disadvantages to the generation

means analysis:

1. An estimate of heritability cannot be obtained; and

2. Genetic advances cannot be predicted because genetic

variances are not available.

Although these disadvantages are inherent to the generation means analysis, they are of

little consequence since additional calculations can be made to obtain the variances

necessary to estimate heritability.

Generation means analysis, as described by Hayman (31), has been used

extensively in other crops such as corn (21,34,70), cotton (Gossypium hirsutum L.)

(55,56), and wheat (8). In most cases, Gamble's (21) notation is used in conjunction

with Hayman's (31) methodology. Generation means analysis apparently has not been

used previously in soybean to analyze quantitative traits. Because of the advantages of








44

this analysis, it was chosen to determine the mode of inheritance of resistance to soybean

looper.

Populations used were derived from the cross D86-3429 x Braxton (Figure 5).

A minimum of six populations are required in a generation means analysis with a six

parameter model, but this experiment was expanded to include nine populations. The

number of individuals comprising each generation are listed in Table 2. To determine

the inheritance of resistance to soybean looper in soybean according to Hayman's (31)

model, programs were written in SAS to analyze the data by cage (see Appendix C) and

combined over cages where cages were treated as blocks (see Appendix D). Using

Hayman's methodology and Gamble's (21) notation, the models for the generation means

analysis were as follows:

P, = m + a d/2 + aa ad + dd/4
P2 = m a d/2 + aa + ad + dd/4
F, = m + d/2 + dd/4
F2 = m
BC, = m + a/2 + aa/4
BC = m a/2 + aa/4
F3 = m d/4 + dd/16
BSi = m + a/2 d/4 + aa/4 ad/4 + dd/16
BS2 = m a/2 d/4 + aa/4 + ad/4 + dd/16


where m = the overall mean, and a, d, aa, ad, and dd represent the additive, dominance,

additive x additive, additive x dominance, and dominance x dominance genetic effects,

respectively.

Each SAS program fitted two regression models which were set up using matrix

notation according to the procedures outlined by Jennings et al. (34). The first












D86-3429 (P)




x




BCI


BS
BS,


Braxton (P2)




x




BC2




BS2


the genetic


Figure 5. Mating scheme to derive populations required to estimate
effects of resistance to the soybean looper in soybean.










Table 2. Populations used for generation means analysis of the inheritance of resistance to
the soybean looper.
Row Plants Total
Population Gen. Number per Rowt Plants
D86-3429 P1 1 11 10 110
Braxton P, 12-22 10 110
D86-3429 x Braxton F, 23 -35 1 13
D86-3429 x Braxton F2 36 85 1 500
D86-3429 x Braxton F3 86 210 10 1250
D86-3429(2) x Braxton BCi 211 -224 1 14
D86-3429(2) x Braxton BS, 225- 249 10 250
Braxton(2) x D86-3429 BC2 250 255 1 6
Braxton(2) x D86-3429 BS2 256 271 10 160
D75-10169 272 274 10 30
PI 229358 275 277 10 30
tRows were 2.5 cm long except where there was only 1 plant/row.
tInsect-resistant germplasm lines used as checks.








47

regression model, defined as Model 1, consisted of the three parameters m, a, and d.

The second regression model, defined as Model 2, consisted of the epistatic effects, aa,

ad, and dd, in addition to the parameters in Model 1. Model 2 is used only if a

significant additive or dominant effect is detected and to determine if significant epistatic

effects exist that are contributing to the significance in Model 1. To adjust for the

unequal population sizes comprising each generation, the models were weighted using

reciprocals of the standard errors of the generation means as suggested by Rowe and

Alexander (65).

In addition, Powers' partitioning method of genetic analysis (49,61) was

incorporated to test proposed genetic models. Initially the frequency distributions of the

PI, P2, FI, and F2 generations were tested for normality by testing the observed versus

theoretical distribution with a X2 test following the procedure previously outlined.

Frequency distributions (observed and theoretical) of the F2 population were converted

to percentages in each class in the following manner:


%(in each class) = n,/n..


where ni was the number of plants in each class and n.. was the total number of plants

in the F2 population. After conversion, the observed distribution of the F2 population

was examined for modes. Finally, the F2 population of each cage was tested for

goodness of fit to a two-gene additive model. The proposed genetic ratio for the model

was as follows: 1:2:1:2:4:2:1:2:1. To test the model, the ratio was simplified to 1:14:1

by combining the defoliation classes in the two cages as follows: 2-4, 5-8, and 9-10 in








48
cage 1, and 2-3, 4-8, and 9-10 in cage 2. The end classes were defined by a distinct

transition between defoliation class in the distribution of the F2 population of each cage.

The homozygous populations D86-3429, Braxton, D75-10169, and PI 229358

were used to determine if there was a significant location effect within each cage. An

analysis of variance was performed on each population to obtain the row and plant within

row mean squares. Estimates for plant to plant variability within a row (o2) and the

additional variability of plants in different rows (or) were obtained from the expected

values of the mean squares. Also, F-tests for significant row effects (Ho:o2r=0) were

performed on each of the homozygous generations within each cage.

An estimate for the number of genes (n) involved in resistance to soybean looper

was obtained for the progeny in each cage by the formula:


n = (iPi iP2)2/8[(oF2) ((v1o2Pi + v2P2 + v3o2F)/v1 + v2+ v)]


In the formula, iP, and xP2, were the mean defoliation ratings for the parents, and o0F2,

02P1, 0P2, and o2Fi were the variances of the respective generations. This formula is

a modification of the formula presented by Poehlman (60). In the modification, o2F, is

replaced with (vo2P+v22P2 +v 32F,)/(v+ v2+v3). This modification was used in order

to better estimate the environmental variance, since o2Fi, a2P1, and o2P2 were essentially

estimates of environmental variation. The degrees of freedom (v,, v2, and v3) for the

associated generations were employed to lessen population size effects on the estimate

of environmental variance. This method of estimating genes assumes that the genes have

equal effects without significant dominant or epistatic effects.








49
Heritability of insect resistance, in the broad sense, was estimated for the progeny

in each cage from the formula:


H = (Vo/Vp) x 100%


where Vo was the genetic variance and Vp was the phenotypic variance. The genetic

variance was:


V, = VP VE


where VE was the environmental variance. The VE was estimated by obtaining the

weighted average of the variances of the F, population and the populations of the parents

P, and P2:


(vor2P, + v2o2P2 + v32F,)/(v, +v2+v,)


In the formula the respective degrees of freedom v were used to weight the variances of

each population. The Vp was estimated by utilizing the variance of the F2 population.

Results and Discussion

1990

The defoliation data of the PI, P2, Fi, and F2 populations from the preliminary

cage study are presented in Table 3. Also presented in this table are the defoliation

ratings of PI 229358 and D75-10169, from which D86-3429 derived its resistance to

soybean looper. D86-3429 was visibly more resistant than the original source of










Table 3. Ratings of leaf feeding by soybean looper on soybean parents, their F, and F2
progeny, and the germplasm lines PI 229358 and D75-10169 in the field cage-1990.
Upper class limits of leaf feeding ratings
1 2 3 4 5 6 7 8 9 n
D86-3429(PI) 4 60 64 11 2.6 139
Braxton(P2) 1 32 48 38 19 1 6.3 139
PI x P2(F) 5 3.0 5
P2 X PI(Fi) 1 4 3.8 5
PI x P2(F2) 1 33 78 154 39 6 2 1 3.7 314
PI 229358 21 25 1 1 3.6 48
D75-10169 4 34 18 3.3 56
tl = 0 to 10% defoliation, 2 = 11 to 20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to
50%, 6 = 51 to 60%, 7 = 61 to 70%, 8 = 71 to 80%, 9 = 81 to 90%, 10 = 91 to
100%. None of the plants received a rating of 10.








51
resistance, PI 229358. This suggests that additional genes for resistance might be

contributing to the increased level of resistance.

Resistance to soybean looper was interpreted as being quantitatively inherited

based on the defoliation data of the Pi, P2, FI, and F2 populations (Table 3). Distribution

of the F2 population (n=314) was continuous, with a mean defoliation rating of 3.7

which was slightly less than the parental midpoint of 4.5. Although a high number of

individuals fell in the fourth rating class (31 to 40% defoliation), the F2 population

appeared to have a normal distribution. The PI, P2, and F2 population distributions were

tested for normality using the procedure outlined by Leonard et al. (49) (see Appendix

E). Both parental populations had low X2 values (Table 4) indicating they were normally

distributed. The frequency distribution of the F2 ratings failed the normal distribution

test (Table 4). Rating class 4 had too many plants, whereas classes 3 and 5 had too few

(Table 3 and Appendix E).

Although the defoliation ratings of the F2 population did not fit a normal curve

the distribution was continuous with only one mode. Therefore use of a method of data

analysis for quantitative traits seemed appropriate in order to define the mode of

inheritance of soybean looper resistance. A generation means analysis (31,34) was

chosen and additional crosses were made to supply the generations necessary for the

experiment.

1991

Larvae were allowed to feed until susceptible Centennial was 95% defoliated.

Plants were found in all rating classes except class one. All populations receiving larger










Table 4. Goodness of fit test for normality of soybean looper defoliation for
populations of D86-3429, Braxton, and D86-3429 x Braxton F2 populations in the 1990
cage study.
Population Degrees of Freedom X2 P
D86-3429 3 3.6628 0.50- 0.25
Braxton 5 7.1877 0.50-0.25
F2 7 22.3204 < 0.005
t Test follows the procedure of Leonard et al. (49).








53

mean defoliation scores in 1991 (Table 5) than in 1990 (Table 3), when larvae were

killed when Centennial was only 85% defoliated.

In 1991 the AV rating method was substituted for the WP method used in 1990.

The AV method accounted for feeding preferences of soybean looper (24). It was

expected to yield a better estimate of defoliation than the WP method, because each score

of the AV method was the average of three observations per plant, whereas the WP

method was based on a single observation. The AL method was too time consuming to

be used in an experiment of this magnitude with the resources available.

Ratings of defoliation by soybean looper on the parents, Pi and P2, and their F,,

F2, and F3 progenies grown in cages 1 and 2 are shown in Table 5. The F2 and F3

progenies, in both cages, exhibited a normal distribution for ratings of defoliation. In

both cages the F, population exhibited a bimodal distribution instead of the expected

single mode. All of the FI plants exhibited purple flowers and tawny pubescence, ruling

out the possibility that the bimodal distribution resulted from selfs in D86-3429.

The Fi defoliation data distribution was expected to exhibit a narrow range and

single mode intermediate to the two parental modes. Instead the distribution had a wide

range and was bimodal in both cages (Table 5). Since the Fi data did not fit expected

distributions, it was suspected that there was a location effect within each cage interfering

with the defoliation ratings. The four homozygous populations (D86-3429, Braxton,

D75-10169, and PI 229358) included in the study were tested to determine if there was

a significant location effect within each cage. A significant location effect was detected

for the populations of D86-3429, Braxton, and D75-10169 in each cage (Table 6).










Table 5. Ratings of leaf feeding by soybean looper on soybean parents, their F1,
BSI, BC2, and BS2 progeny, and the germplasm lines PI 229358 and D75-10169


F2, F3, BCI,
in the field


cages-1991.
Upper Class Limits of Leaf feeding ratings
2 3 4 5 6 7 8 9 10 it n


D86-3429(P,)
Braxton(P2)
P, x P2(F1)
PI x P2(F)
P x P2(F3)
P, x (PI x P2)(BCI)
P1 x (PI x P2)(BS,)
P2 x (PI x P2)(BC2)
P2 x (PI x P2)(BS2)
PI 229358
D75-10169

D86-3429(P,)
Braxton(P2)
P1 X P2(F,)
Pl X P2(F2
P1 x P2(F3)
PI x (P1 x P2)(BCi)
PI x (PI x P)(BSI)
P2 x (PI x P2)(BC2)
P2 x (P, x P2)(BS2)
PI 229358
D75-10169


Cage 1
3 30 44 19 2
2
1 2 2
1 8 31 58 102
7 37 118 194 232


5
2 34



1
2 7


8 32 43
1
1 2 1
2 17 43


28
2
5 23


103
5
59


1
3
3 9


1
19

10
11
2
Ca2e 2


2
58 82
89 238
3 2
51 37

5 12
16 8
9 6


18
6
111
201
1
8

13
1


5
3
120
273
2
16

17
1


3.5
30 40 4 8.0
5.8
67 36 4 6.2
148 131 18 6.0
2 4.4
6 4.2
3 1 7.9
52 45 9 7.9
5.1
3.6


93 30 5
187 114 21

7 1
2
33 43 11


tBased on the mean of three observations per plant.


1 = 0 to 10% defoliation, 2 = 11 to


20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to 50%, 6 = 51 to 60%, 7 = 61 to 70%, 8 =
71 to 80%, 9 = 81 to 90%, 10 = 91 to 100%. None of the plants received a rating of 1.
:Derived from the mean defoliation of all plants without regard of the upper class limits.


98
94
11
418
1086
14
194
4
130
21
28

97
94
12
450
1153
14
199
2
122
28
27










Table 6. Estimates of variance components obtained from an analysis of the homogeneous
populations, D86-3429, Braxton, D75-10169, PI 229358, for a location effect due to
feeding by soybean looper in field cages in 1991.
F
Population do2,* 2 + 2,1' Ho:e,=0


Cage 1
D86-3429 (PI) 0.31 0.72 12.55"'
Braxton (P) 0.25 0.84 20.78"*
D75-10169 (BD) 0.29 1.50 40.29"*
PI 229358 (D) 0.58 0.96 5.37'
Cage 2
D86-3429 (P,) 0.37 0.54 5.03**
Braxton (P) 0.69 0.80 2.35'
D75-10169 (BD) 0.42 0.94 12.03'"
PI 229358 (D) 0.39 0.41 1.41
Significant at the 0.05 and 0.001 probability levels respectively.
t Based on defoliation ratings made with a 1 to 10 scale divided into 10% increments with
a 10 reflecting 91-100% damage.
* Variance of plants within same row.
* Variance of plants among rows.








56

The location effect for the PI 229358 population was significant (p < 0.05, Table 6) in

cage 1, but was nonsignificant in cage 2. Although the population size of PI 229358 is

approximately the same for each cage, different randomization plans would produce

different results. It is apparent that the defoliation ratings were affected by location

within the cages.

The mean ratings for soybean looper defoliation and their standard errors for the

nine generations grown in each cage are presented in Table 7. In addition to the

observed mean ratings for defoliation, the predicted means determined from relationships

described by Mather and Jinks (51) are also presented. In all instances except the BC2

generation in cage 2, the predicted means were within + one standard deviation of the

observed means in each cage. The observed mean for the BC2 generation in each cage

was larger than the predicted mean. This deviation from the predicted mean could be

associated with the small sample size available for defoliation ratings. Variances of the

generations from each cage are presented in Table 7, and are used later in Mather's

scaling test (51) and to determine heritability.

The results of Mather's scaling test (51) for the cross D86-3429 x Braxton are

shown in Table 8. According to Mather and Jinks (51), if the values of A, B, C, and

D do not differ from zero, then additive gene effects are indicated. The significance of

any of these scales indicates the presence of epistasis. In cage 1, nonsignificant values

for A, B, C, and D, indicated that an additive-dominance model was adequate. Cage 2

data yielded similar t-tests for A, C, and D, supporting the conclusions from cage 1.

The B scale test could not be applied to cage 2 since the BC2 population had a variance










Table 7. Means, standard errors, and variances of soybean looper defoliation ratings for
parental lines and seven populations of soybean arising from the cross D86-3429 x
Braxton grown in field cages at Stoneville, MS in 1991.
Mean
-------- :i


Population Gen.


D86-3429
Braxton
P1 XP2
P x P2
Pi x P2
Pi x (PI
Pi X (PI
P2 X (P1
P2 X (PI:


P1
P2
F1
F2
F3
BCI
BSI
BC2
BS2


D86-3429
Braxton
P x P2
PI XP2
P1 XP2
Pi x (PI x P2)
P, x (PI x P2)
P2 X (PI X P,)
P2 x (P1 x P2)
tBased on relationships
fStandard error


P,
P2
F,
F2
F3
BCi
BS,
BC2
BS2
described


Number Actual
Cage 1
98 3.49
94 7.97
11 5.76
418 6.19
1086 6.02
14 4.36
194 4.22
4 7.92
130 7.89
Cage 2
97 3.34
94 8.28
12 5.42
450 6.13
1153 6.18
14 4.45
199 4.52
2 9.33
122 7.69


Predt S.E.t a2


5.73
5.75
5.96
4.63
4.62
6.87
6.86




5.81
5.62
5.97
4.38
4.48
6.85
6.95


0.084
0.092
0.427
0.072
0.052
0.524
0.087
0.394
0.090

0.074
0.092
0.617
0.074
0.046
0.376
0.093
0.000
0.117


by Mather and Jinks, 1971.


0.685
0.797
2.002
2.160
2.916
3.837
1.452
0.620
1.055

0.528
0.788
4.568
2.490
2.478
1.976
1.708
0.000
1.655








58

Table 8. Mather's scaling test applied to the cross D86-3429 x Braxton to test adequacy of
additive-dominance model for resistance to soybean looper.

Test Cage 1 Cage 2

tA -0.125 0.039
tB 0.918
tc 0.268 0.267
tD 0.157 0.273








59
of zero. Since C was not significantly different from zero in either cage (Table 8), this

indicated that generation means depended only on the additive gene effects and no

epistasis was present.

The results of the generation means analysis of each cage are reported in Table

9. Table 10 reports the results of the combined analysis where the cage effects are

treated as blocks. Significant negative additive effects were detected with Model 1,

which measures only additive and dominant effects, for both Cages 1 and 2 (p < 0.06

and p < 0.02, respectively; Table 9) and the combined cages (p < 0.001, Table 10).

The sign of this effect is dependent upon the parent designated as P,. The sign of the

effect is a function of the fact that an inverted scale was used and also a reflection of the

relationship between the mid-parent and the means of the F,, F2, and F3 generations

indicating which parent was contributing to the additive variation (51). In Cage 1 (Table

7) the means of the Fi, F,, and F3 generations were between the mid-parent (5.73) and

P2. The means of the F, and F3 in Cage 2 were skewed away from the mid-parent (5.81)

in the same direction, but the F, mean was oriented towards the other parent. Although

the Fi's mean was on the opposite side, it was within one standard deviation of the mean.

All of the progeny means skewed towards P2, in addition to a significant negative

additive effect indicated that Braxton might be contributing to the additive variation.

Since it has been reported that Braxton does have some level of resistance (64), it stands

to reason that it would contribute to the additive effect.

The full model, Model 2, was then fitted to estimate the epistatic effects as well

as the remaining additive and dominant effects. None of the epistatic (aa, ad, and dd),










Table 9. Estimates of the additive, dominant, and epistatic effects in the generation
means for defoliation by soybean looper in the nine populations of the D86-3429 x
Braxton material grown in the field cages at Stoneville, MS in 1991.
Standard
Effects Estimate error t value
Cage 1
Model 1
F2 mean (m) 5.503 0.655 8.399"*
Additive (a) -1.830 0.805 -2.273t
Dominance (d) -1.287 2.071 -0.621
Model 2
F2 mean (m) 5.839 0.775 7.538"
Additive (a) 0.859 2.525 0.340
Dominance (d) -2.639 3.103 -0.851
Additive x Additive (aa) -2.182 2.642 -0.826
Additive x Dominance (ad) 3.212 2.737 1.174
Dominance x Dominance (dd) 3.484 9.370 0.372
Cage 2
Model 1
F2 mean (m) 5.387 0.575 9.374*"
Additive (a) -2.206 0.667 -3.308'
Dominance (d) -1.385 1.796 -0.771
Model 2
F2 mean (m) 5.613 0.590 9.507"
Additive (a) -0.031 1.929 -0.016
Dominance (d) -2.673 2.525 -1.059
Additive x Additive (aa) -2.191 1.953 -1.122
Additive x Dominance (ad) 2.534 2.063 1.228
Dominance x Dominance (dd) 4.302 7.742 0.556
,","" Significant at the 0.05, 0.01, and 0.001 probability
levels,respectively.
t Significant at the 0.06 probability level.










Table 10. Estimates of the additive, dominant, and epistatic effects in the generation
means for defoliation by soybean looper in the nine populations of the D86-3429 x
Braxton material grown in the field cages at Stoneville, MS in 1991.
Standard
Effects Estimate error t value
Cages 1 and 2 Combined
Model 1
F2 mean (m) 5.414 0.503 10.760"'
Additive (a) -2.020 0.487 -4.150"*
Dominance (d) -1.327 1.281 -1.040
Model 2
F2 mean (m) 5.690 0.418 13.630'"
Additive (a) 0.424 1.165 0.360
Dominance (d) -2.669 1.468 -1.820
Additive x Additive (aa) -2.195 1.197 -1.830
Additive x Dominance (ad) 2.883 1.254 2.300'
Dominance x Dominance (dd) 3.900 4.464 0.870
*,*" Significant at the 0.05 and 0.001 probability levels,respectively.








62
additive, or dominant effects were significantly different from zero in either cage.

Although unexpected, ad was significant (p < 0.05) in the combined analysis and neither

the additive or dominant effects were significant. It is interesting and informative that

ad effect is significant. However, it introduces a dimension that cannot be examined

further with currently available data. In Model 1 the epistatic effects are included in the

additive effects. For this reason, Model 1 will be assumed to treat the data adequately.

With an indication that inheritance of resistance to soybean looper was controlled

by an additive genetic system, the data from each cage were tested to fit a two-gene

model. Since no modes were evident in the F2 or F3 populations in either cage (Table

5), Power's partitioning analysis was employed to analyze the data and test the predicted

gene models following the examples of Deren and Quesenberry (12) and Sage and de

Isturiz (67).

Initially, the frequency distributions of the defoliation data from the parental, F,,

and F2 populations from each cage were tested for normality (Table 11). As was seen

in 1990, both parental populations had low X2 values which indicated that these

populations were normally distributed about a central mean. In cage 1, the Fi population

did not have a normal distribution, but the F2 population's distribution did fit the normal

curve. The opposite situation occurred with the Fi and F2 populations in cage 2. The

differences between the two cages with respect to the F, and F2 populations may be due

to a location effect present within the cages.

In the hypothesized two-gene additive model the middle classes were combined

resulting in three classes for the model to be tested. Partitioning the middle classes










Table 11. Goodness of fit test for normality of soybean looper defoliation data from 1991
cage study.

Population Degrees of Freedom X2 P

Caege 1


1.519

7.767

10.667

6.528


0.211

4.584

4.000

26.599


0.90 0.75

0.25- 0.10

0.025 0.010

0.75 0.50



0.990 0.975

0.500 0.250

0.75 0.50

0.005 >


D86-3429


Braxton

F,

F2


Cage 2


D86-3429


Braxton

F1

F2








64
would have been arbitrary, whereas the end class ratios were easily manipulated to test

the model. The F2 data for each cage were tested for goodness of fit to the ratio of

1:14:1 for the two-gene model. The chi-square probabilities were as follows: cage 1,

0.005 > p; and cage 2, 0.10 > p > 0.05. These probabilities are reflective of the test

for normality of the F, defoliation data in both cages as shown in Table 11. The higher

the probability of a normal distribution of the F2 defoliation data, the lower the

probability that the F2 defoliation data fit the two-gene model. Neither F2 population fit

the two-gene model well. The three-gene model gave a better fit. However, because

there is good reason to believe that the scores assigned in this experiment should not be

assumed to correspond to genotype effects, the Power's partitioning analysis probably has

no utility in these analyses.

An estimate for the number of genes, n, in this cross contributing to the

expression of resistance to soybean looper was obtained for both cages using a

modification of the formula presented by Poehlman (60). The number of genes

contributing to the expression of resistance to the soybean looper in the cross D86-3429

x Braxton was estimated at n = 1.87 and n = 1.91, cages 1 and 2, respectively.

Broad sense heritability was estimated for resistance to soybean looper in the

progeny from both cages. Although it has already been concluded that resistance to

soybean looper is controlled by additive gene action, narrow-sense heritabilities could not

be calculated because of the size and variances of the BCI and BC, populations in each

cage. Although broad sense heritabilities were calculated, it is assumed that these

heritabilities are attributed to only the additive genetic variance and are in fact narrow-








65
sense heritabilities. Since the F, generations in both cages exhibited a bimodal

distribution for defoliation scores, the environmental variance was estimated by the

following formula:

(vP,2P, + vpP,2 + vF,2Fl)/(vp, + v,2 + VF)

as was done in estimating the number of genes. This gives a better estimate of the

environmental variance than just using the Fi because the number of individuals

comprising the parental generations were relatively large (Table 5). Also, parents are

genetically uniform and should only exhibit variability due to the environment. The

heritabilities were estimated at 62.3% for cage 1 and 64.3% for cage 2. As has been

demonstrated in all previous analyses the results from the two cage studies are in

agreement for the estimate of broad sense heritability. Based on the high heritabilities

for resistance to soybean looper, it is apparent that the phenotypes expressed in this

population were determined by the genetic variation more than by the environmental

variation.

Although the estimates for both cages are in agreement, the estimate is probably

an underestimation of the total number of genes contributing to soybean looper resistance

because these data only estimate the number of genes controlling soybean looper

resistance by which D86-3429 and Braxton differ. In both cages, Braxton exhibited a

low level of resistance with 80% defoliation, whereas Centennial was almost completely

defoliated. When compared to Centennial, which is known for its high susceptibility

(47), Braxton has repeatedly demonstrated 15 to 20% less defoliation. Since the

inception of this study, resistance of Braxton to foliar-feeding insects has been








66
documented by Rowan et al. (64) in a study to compare the resistance of 46

recommended soybean cultivars in Maturity Groups V, VI, VII, and VIII. In that study,

Braxton was cited as being one of the most resistant cultivars in Maturity Group VII.

As in 1990, the insect-resistant germplasm lines, PI229358 and D75-10169, were

included in both cages to provide additional checks and to document the level of

resistance in the resistant parent, D86-3429. In both cages, D86-3429 was visibly and

statistically (Table 12) more resistant than the original source of resistance, PI 229358.

D86-3429 also exhibited more resistance than did D75-10169 in both cages, although it

was significantly different at p = 0.01 (LSD) only in cage 2 (Table 12). This suggests

that during the five cycles of crossing and selection that occurred in the development of

D86-3429 from PI 229358, additional genes were gained from 'Tracy-M', a parent in the

fourth cycle. Tracy-M has subsequently been identified as being measurably resistant to

foliar-feeding insects (47).

Summary and Conclusions

This research was conducted to determine the mode of inheritance of resistance

to soybean looper in soybean. The results from Mather's scaling test (51) applied to the

data from each cage study indicated that generation means depend only on the additive

gene effects for resistance to soybean looper. Utilizing Hayman's (31) methodology in

terms of the generation-means analysis to analyze the data from each cage separately,

only the additive genetic effects were significant. When the data from the two cages

were combined for analysis the ad epistatic effect was significant (t = 2.30, p < 0.005).

However, the primary effect is assumed to be additive based on the results of the analysis










Table 12. Average defoliation scores in soybean populations resistant to soybean looper in
1991 cage study.

Population n Meant

Cage 1

D86-3429 98 3.49 a

D75-10169 28 3.64 a

PI 229358 21 5.13 b

Cage 2

D86-3429 97 3.34 a

D75-10169 27 4.35 b

PI 229358 28 5.42 b

t Means followed by the same letter are not significantly different at the p = 0.01 (LSD).








68
of the data from cage 1 (t = -2.273, p < 0.06), cage 2 (t = -3.308, p < 0.05), and

combined (t = -4.15, p < 0.001), as was indicated by the scaling test. When Power's

partition analysis (49,61) was applied to the data from each cages the F2 data did not fit

a two-gene additive model for resistance to soybean looper.

A modification of Poehlman's (60) formula for determination of gene number was

used to estimate the number of genes controlling soybean looper resistance by which the

parents D86-3429 and Braxton differ. Two genes were estimated to contribute to the

expression of resistance to soybean looper in this cross. Since the formula used only

estimates the number of genes in which the two parents, the actual number of genes for

resistance to soybean looper is probably greater than two. It would take further research

with PI 229358 and/or its derived breeding lines to determine exactly how many genes

are controlling resistance to soybean looper. Also, the information generated from this

research could be integrated with the results of research involving other sources of

resistance (36,45,54,72,83) to determine the number of genes that not only control

resistance to soybean looper, but other insect species as well.

Heritability was estimated to be 63% for resistance to soybean looper. Broad

sense heritabilities were calculated because the size and variances of the BC, and BC2

populations were insufficient, thus prohibiting calculation of narrow-sense heritability.

The previous analyses suggests that resistance to soybean looper in this cross is

controlled by two genes which behave additively, therefore it can be assumed that this

heritability is due only to the additive genetic variance. Since the genetic variation








69
influences resistance to soybean looper more than environmental variation, a breeder

should be able to make selections in the F2 population with confidence.

By utilizing a population from a cross between a resistant breeding line and a

highly susceptible breeding line, it should be possible to determine if there are only three

genes for resistance to soybean looper in breeding lines derived from PI 229358. The

single seed descent method of breeding (60) could be used to determine the number of

genes involved by advancing a large population to the F6 generation and evaluating

replicated hills for reaction to soybean looper. For example, assuming three segregating

gene pairs, 91 % of the F6 lines should be homozygous (6) for their respective genotypes,

thus allowing for replications of each genotype. Replicating the F6 lines, as well as their

parents, increases the precision and accuracy of the measured reaction to soybean looper

for each genotype. Examination of results from this type of experiment should indicate

whether or not there are only three genes controlling resistance to soybean looper.

Another way to determine if there are only three genes for resistance to soybean

looper in PI 229358 derived genotypes would be to utilize restriction fragment length

polymorphism (RFLP) markers. RFLP markers have been widely used in other plant

species to construct linkage maps and to place genes that control qualitative and

quantitative traits onto these maps. Recently, RFLP markers have been used to map

phytophthora resistance loci in soybean (13). This methodology could be used to map

the genes and possibly determine any linkage groups that may be associated with the

genes controlling resistance to soybean looper.








70
Since the results of this study have shown that resistance to soybean looper was

a quantitative trait that was additive, it should be possible to increase the level of

resistance in future breeding lines. This might be accomplished by utilizing new sources

of resistance which have additional genes for insect resistance. Another method of

increasing the level of resistance to soybean looper would be by interspecific

hybridization which would allow genes for resistance in other species to be incorporated

into the soybean plant.

To fully exploit the benefits of insect resistant breeding lines, it will become

necessary to understand the mechanisms of resistance that the plants possess. Depending

on the source of resistance, the mechanism of resistance may be different in the various

sources. Several researchers (39,53,63,73,81) have already conducted studies on the

mechanisms of resistance, but further research is needed. As new technology becomes

available, the mysteries of the mechanisms of insect resistance and the genes that control

their expression are closer to being solved.













CHAPTER 5
SUMMARY AND CONCLUSIONS



Soybean is grown worldwide primarily for its oil and meal products, and to a

lesser degree the whole bean product. As in any crop, soybean is exposed to biotic and

abiotic yield-reducing factors such as diseases, nematodes, insects, and drought

throughout the growing season. To produce a profitable crop a grower must modify the

effects of these yield-reducing factors by one or more of the methods described by

Metcalf and Luckmann (50). Environmentally, the best approach to any pest problem

would be to utilize genes for resistance in available germplasm to develop resistant

cultivars (82).

Soybean looper is a major pest to soybean in many areas of the southeastern

United States. The soybean looper, more so than any other arthropod pest, has become

increasingly resistant to available insecticides (10,18,48,58,60) making it difficult to

control. By utilizing soybean germplasm lines that possess genes for resistance to foliar-

feeding insects, it is possible to increase the host plant responses for control as has been

demonstrated by the development of the insect-resistant cultivars Lamar (28) and

Crockett (5).

The objectives of this study were: (i) to evaluate three methods of rating for

defoliation of soybean by soybean looper, and (ii) to determine the inheritance of








72
resistance in soybean to the soybean looper. The latter was the most important objective

since knowledge of inheritance of a particular trait increases the efficiency for

incorporating genes for that trait into an agronomically desirable plant type. Rating

methodology can affect the accuracy of classification and therefore was evaluated

separately to offer guidelines for future insect defoliation studies.

Evaluation of Rating Methods

The three methods of defoliation assessment evaluated in 1990 and 1991 were the

whole plant visual rating (WP), the partitioned plant average visual rating (AV), and

measured average leaf area of the partitioned plant (AL). The WP and AV were visual

assessments of defoliation based on a 1-to-10 scale divided into 10% increments with a

score of 1 signifying defoliation of 0 to 10% and a score of 10 signifying defoliation of

91 to 100%. The WP method was based on one observation per plant, whereas the AV

method was based on the mean of three observations per plants. The absolute method

(AL) utilized a LI-COR LI-3000 leaf area meter to measure the area of three trifoliolates,

one each from the top, middle, and bottom of the plant.

In 1990 no absolute method comparisons were possible, however the WP and AV

were highly correlated (r=0.65, p < 0.001). Adjustments in the method of calculating

percent defoliation by AL for 1991 allowed for direct comparisons with the WP and AV

methods. All correlations (WP vs. AV, WP vs. AL, and AV vs. AL) were significant

at the p < 0.001 level whether the cages were examined separately or pooled. The

results of these correlations indicate that the evaluator could obtain similar estimates of

defoliation regardless of which of the three methods is utilized. These findings are








73
similar to the results of another relative versus absolute estimation of defoliation study

conducted in 1985 by G.R. Bowers (G.R. Bowers, 1992, personal communication).

A paired difference comparison was performed to test the equality of the

defoliation assessment between the relative methods in 1990 and among the three

methods in 1991. The results indicated that estimates of defoliation by the two relative

methods in 1990 and by all three in 1991 were significantly different. With the relative

methods, WP and AV, the evaluator estimated somewhat similar amounts of defoliation

across all plants when compared with each other, even though the difference between the

two methods was highly significant. However, in comparisons with an absolute method,

as in 1991, it becomes apparent that the evaluator's estimates of defoliation with the WP

method were closer to the actual than were the estimates of the AV method. These

findings were validated with McNemar's test for concordance.

The relative variation (RV) was calculated for each method as a measure of

precision. In 1990 the precision of the WP and AV methods were essentially equal.

Similar results were obtained in 1991 by all three methods. Both years the precision of

the methods was essentially equal and well below 10, the value that Pedigo states as

being desirable in research. These results indicate that either relative method, WP or

AV, would be suitable for estimating defoliation by insects.

Based on the results of these studies it is apparent that relative estimates of

defoliation can be used in genetic studies or routine screenings of advanced breeding

lines, in lieu of an absolute estimate. Since the two relative methods yielded similar

results it would advantageous to use the simplest and most economical method for








74
making estimates of defoliation. With increased experience in visual estimates of

defoliation, researchers should be able to estimate defoliation with reliability and

repeatability.

Inheritance Study

This study was undertaken to determine the mode of inheritance of resistance to

the soybean looper in soybean. The study consisted of two years of field cage

evaluations of progeny derived from the cross D86-3429 x Braxton. The results of the

preliminary study in 1990 suggested that the genes controlling resistance to soybean

looper might be inherited quantitatively. The 1991 experiment was designed so that

inheritance of resistance to soybean looper could be examined quantitatively as well as

qualitatively. Also, the 1991 experiment was duplicated in two field cages so that the

experiment would be repeated in two separate environments to strengthen the results and

conclusions of the study.

In the 1991 study, estimates of defoliation were made for each plant using the AV

method. Although this method was more time consuming than the WP method, it was

chosen because its estimate of defoliation was based on three observations per plant

rather than one observation. After assessing each plant in each cage for defoliation by

the soybean looper the data generated were analyzed by several methods of genetic

analysis.

Mather's scaling test (51) was applied to the defoliation data from each cage study

and it was determined that generation means depend only on the additive gene effects and

no epistasis was present. Utilizing Hayman's (31) methodology in terms of the








75
generation-means analysis to analyze the data from each cage separately, only the

additive genetic effects were significant. When the data from the two cages were

combined for analysis the ad epistatic effect was significant (t = 2.30, p < 0.005).

However, the primary effect is assumed to be additive based on the results of the analysis

of the data from cage 1 (t = -2.273, p < 0.06), cage 2 (t = -3.308, p < 0.05), and

combined (t = -4.15, p < 0.001), as was indicated by the scaling test. Applying

Power's partitioning analysis (49,61) to the data from each cage it was determined that

in this cross the F, data did not fit the two-gene additive model for resistance to soybean

looper. There is good reason to believe that the defoliation scores assigned in this

experiment should not be assumed to correspond to genotype effects, therefore the

Power's partitioning analysis has no utility in these analyses.

Based on a modification of the formula presented by Poehlman (60), it was

determined that two genes contribute to the expression of resistance to soybean looper

in this cross. However this formula only estimates the number of genes for resistance

to soybean looper by which the two parents differ. In actuality the number of genes for

resistance to soybean looper is probably greater than two, but it would take further

research to determine this.

Broad sense heritability based on the defoliation data was estimated to be 63% for

resistance to soybean looper. Since the previous analyses suggest that resistance to the

soybean looper is controlled by two genes which behave additively, it can be assumed

that this heritability is due only to the additive genetic variation. A heritability of this








76
magnitude suggests that a breeder should be able to make progress by selecting for

resistance to soybean looper in early generations.

This research could be continued with breeding lines derived from PI 229358 or

by incorporating additional sources of resistance, such as PI 227687, PI 171451, PI

417061, or PI 507301 into agronomically adapted breeding lines. These elite genotypes

could then be cross-pollinated with a highly susceptible cultivar, such as Centennial, to

develop large F2 populations for future evaluations. In order to determine the number

of genes controlling resistance to the soybean looper and/or other species, these F2

populations could be screened by conventional methods or some of the newer methods,

such as RFLP analysis.

Since resistance to soybean looper in this study was determined to be additive, it

should be possible to increase resistance in future cultivars by pyramiding additional

genes from other sources of resistance. This can be done by conventional plant breeding

techniques by utilizing soybean germplasm lines that show resistance to leaf-feeding

insects and incorporating these genes into adapted cultivars or advanced breeding lines.

The same goals could also be accomplished by any of the new methods of biotechnology

that are currently available.

Additional research will need to be conducted on sources of resistance other than

PI 229358-derived lines to determine how many other genes for resistance are available

to breeders for utilization in their breeding programs. In order to fully exploit the

benefits of insect resistance in a breeding program, it will become necessary to

understand the mechanisms of resistance as well as the underlying genes that code for the








77

mechanism. As new technology becomes available, the mysteries of the mechanisms of

insect resistance will be solved. Ideally, the future promises the development of a

resistant cultivar that has many genes for resistance that code for all three mechanisms

of resistance.































APPENDIX A
LIST OF DEFOLIATION SCORES GENERATED BY WHOLE PLANT VISUAL,
PARTITIONED PLANT VISUAL, AND LEAF AREA OF THE PARTITIONED
PLANT RATING METHODS IN 1990.










List of defoliation scores generated by whole plant visual, partitioned plant visual, and leaf
area of the partitioned plant rating methods.
Whole Avg. Avg.
Observation Plant Plant Visual Leaf Area
1 BBE-5-3 1 2.67 57.55
2 BBE-5-4 2 2.00 66.45
3 BBE-5-5 4 1.67 72.74
4 BBE-5-6 3 2.33 65.45
5 BBE-5-7 3 3.00 81.61
6 BBE-5-8 2 2.33 75.80
7 BBE-5-9 2 2.33 111.43
8 BBE-5-10 3 2.67 66.73
9 BBE-6-1 4 3.00 84.04
10 BBE-6-2 2 2.33 47.86
11 BBE-6-3 2 2.33 66.22
12 BBE-6-4 4 2.33 32.03
13 BBE-6-5 3 2.00 87.55
14 BBE-6-6 3 2.67 86.85
15 BBE-6-7 2 2.33 52.67
16 BBE-6-8 4 3.33 76.31
17 BBE-6-9 4 3.33 71.35
18 BBE-6-10 4 2.00 64.61
19 BBE-7-1 3 4.33 100.27
20 BBE-7-2 3 3.00 71.86
21 BBE-7-3 4 3.33 68.00
22 BBE-7-4 3 2.67 75.74
23 BBE-7-5 4 1.33 61.07
24 BBE-7-6 3 2.00 77.64
25 BBE-7-7 4 2.67 67.08
26 BBE-7-8 4 4.33 74.19
27 BBE-7-9 2 1.67 73.05
28 BBE-7-10 4 2.00 63.85
29 BBE-8-1 3 3.33 58.87
30 BBE-8-2 3 1.00 88.32
31 BBE-8-3 2 2.00 72.49
32 BBE-8-4 4 2.00 70.66
33 BBE-8-5 4 2.67 73.00
34 BBE-8-6 4 2.00 79.00





Observation
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70


Plant
BBE-8-7
BBE-8-8
BBE-8-9
BBE-8-10
BBE-9-1
BBE-9-2
BBE-9-3
BBE-9-4
BBE-9-5
BBE-9-6
BBE-9-7
BBE-9-8
BBE-9-9
BBE-9-10
BBE-10-1
BBE-10-2
BBE-10-3
BBE-10-4
BBE-10-5
BBE-10-6
BBE-10-7
BBE-10-8
BBE-10-9
BBE-10-10
BBE-11-1
BBE-11-2
BBE-11-3
BBE-11-4
BBE-11-5
BBE-11-6
BBE-11-7
BBE-11-8
BBE-11-9
BBE-11-10
BBE-12-1
BBE-12-2


Whole
Plant
3
3
2
4
4
3
5
4
4
3
4
3
3
4
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
2
3
3
2
4
3
3


Avg.
Visual
2.00
1.67
2.00
2.67
4.67
3.67
5.00
5.00
5.33
3.00
2.67
1.67
2.67
3.67
1.33
2.67
3.67
3.00
1.67
2.67
2.00
2.67
1.33
3.33
7.00
4.67
4.00
3.67
4.33
3.33
3.67
5.33
1.67
6.67
4.33
2.33


Avg.
Leaf Area
80.88
94.90
60.38
31.06
70.52
66.67
47.64
69.57
57.64
87.20
74.37
75.70
66.17
74.00
45.08
77.06
63.51
67.74
71.65
70.88
76.32
89.68
52.45
29.95
64.62
75.10
55.21
65.95
62.94
88.50
74.01
63.24
97.20
74.45
62.75
78.93





Observation
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106


Plant
BBE-12-3
BBE-12-4
BBE-12-5
BBE-12-6
BBE-12-7
BBE-12-8
BBE-12-9
BBE-12-10
BBE-13-1
BBE-13-2
BBE-13-3
BBE-13-4
BBE-13-5
BBE-13-6
BBE-13-7
BBE-13-8
BBE-13-9
BBE-13-10
BBE-14-1
BBE-14-2
BBE-14-3
BBE-14-4
BBE-14-5
BBE-14-6
BBE-14-7
BBE-14-8
BBE-14-9
BBE-14-10
BBE-4-1
BBE-4-2
BBE-4-3
BBE-4-4
BBE-4-5
BBE-4-6
BBE-4-7
BBE-4-8


Whole
Plant
3
3
3
2
3
3
3
2
4
2
2
4
3
4
3
3
3
4
3
3
2
2
4
3
4
4
2
5
3
3
4
3
3
5
4
2


Avg.
Visual
2.67
2.00
2.33
2.33
4.33
3.00
4.33
4.00
5.00
1.67
1.67
4.00
5.00
4.00
4.00
2.33
3.00
5.00
4.67
5.00
4.33
2.00
5.00
3.67
6.33
4.67
3.00
5.00
2.67
2.67
3.33
2.00
4.67
5.67
4.33
2.33


Avg.
Leaf Area
72.66
92.53
93.98
106.10
72.10
86.99
98.32
60.94
53.08
91.42
73.57
75.45
46.18
63.52
68.61
91.37
51.28
58.01
57.90
65.62
65.58
83.45
57.62
84.46
58.82
50.44
79.47
42.37
69.64
72.35
74.24
80.57
49.60
60.68
80.55
72.02










Whole Avg. Avg.
Observation Plant Plant Visual Leaf Area
107 BBE-4-9 3 2.67 58.72
108 BBE-4-10 3 5.67 65.01
109 BBE-1-1 5 6.67 77.55
110 BBE-1-2 3 2.33 92.04
111 BBE-1-3 3 2.67 72.01
112 BBE-1-4 4 5.00 63.01
113 BBE-1-5 4 5.67 64.55
114 BBE-1-6 4 6.00 51.66
115 BBE-1-7 4 5.33 82.70
116 BBE-1-8 3 2.67 64.59
117 BBE-1-9 3 4.33 60.43
118 BBE-1-10 4 3.67 56.59
119 BBE-3-1 2 3.33 69.57
120 BBE-3-2 2 3.33 64.69
121 BBE-3-3 3 2.67 69.09
122 BBE-3-4 2 1.67 60.31
123 BBE-3-5 2 3.00 82.66
124 BBE-3-6 2 2.33 68.51
125 BBE-3-7 2 2.33 70.63
126 BBE-3-8 2 2.33 65.65
127 BBE-3-9 3 2.67 80.47
128 BBE-3-10 3 2.33 69.33
129 BBE-2-1 2 2.00 80.21
130 BBE-2-2 2 2.33 81.01
131 BBE-2-3 2 2.00 67.84
132 BBE-2-4 2 2.67 52.42
133 BBE-2-5 2 2.00 73.62
134 BBE-2-6 3 4.00 82.98
135 BBE-2-7 3 2.33 77.75
136 BBE-2-8 2 2.00 69.91
137 BBE-2-9 2 2.00 64.00
138 BBE-2-10 3 3.00 79.92
139 BBE-16-1 4 5.33 51.63
140 BBE-16-2 4 6.33 45.94
141 BBE-16-3 5 6.00 40.06
142 BBE-16-4 4 4.00 62.02





Observation
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178


Plant
BBE-16-5
BBE-16-6
BBE-16-7
BBE-16-8
BBE-16-9
BBE-16-10
BBE-15-1
BBE-15-2
BBE-15-3
BBE-15-4
BBE-15-5
BBE-15-6
BBE-15-7
BBE-15-8
BBE-15-9
BBE-15-10
BBE-17-1
BBE-17-2
BBE-17-3
BBE-17-4
BBE-17-5
BBE-17-6
BBE-17-7
BBE-17-8
BBE-17-9
BBE-17-10
BBE-18-1
BBE- 18-2
BBE-18-3
BBE-18-4
BBE- 18-5
BBE-18-6
BBE-18-7
BBE-18-8
BBE-18-9
BBE-18-10


Whole
Plant
5
6
5
7
4
7
5
3
4
3
3
4
4
3
4
4
4
5
3
4
3
4
3
4
4
4
4
4
5
4
4
5
4
4
4
4


Avg.
Visual
5.67
7.00
4.67
5.67
5.33
7.33
5.33
4.00
4.67
3.67
3.33
5.33
5.00
4.33
4.67
5.67
6.00
6.00
4.00
6.00
4.33
5.33
5.33
4.67
4.33
5.33
5.67
5.33
5.00
5.33
5.33
5.33
5.00
3.00
5.33
5.00


Avg.
Leaf Area
45.71
35.54
75.60
37.47
59.21
62.08
74.96
62.52
64.42
67.52
76.06
50.96
70.10
74.26
64.68
46.88
44.94
39.03
80.14
39.94
52.10
42.98
78.97
68.50
52.05
58.14
56.14
61.43
51.84
54.93
58.81
53.30
75.94
62.40
61.60
78.89










Whole Avg. Avg.
Observation Plant Plant Visual Leaf Area
179 BBE-19-1 5 5.67 56.32
180 BBE-19-2 3 3.33 74.48
181 BBE-19-3 5 4.00 49.36
182 BBE-19-4 5 6.00 58.93
183 BBE-19-5 3 4.67 55.07
184 BBE-19-6 3 4.33 88.80
185 BBE-19-7 5 5.67 50.52
186 BBE-19-8 4 5.67 57.10
187 BBE-19-9 4 6.33 48.33
188 BBE-19-10 4 3.00 74.01
189 BBE-20-1 4 3.67 53.10
190 BBE-20-2 3 2.67 71.80
191 BBE-20-3 4 4.33 63.00
192 BBE-20-4 3 4.00 60.55
193 BBE-20-5 2 2.67 66.57
194 BBE-20-6 4 3.33 69.70
195 BBE-20-7 3 3.33 72.31
196 BBE-20-8 4 3.67 56.56
197 BBE-20-9 3 4.33 64.67
198 BBE-20-10 4 2.67 53.80
Mean (n= 198) 3.32 3.68 67.13
Variance 0.98 2.07 201.92
Std.Dev. 0.99 1.44 14.21
t Based on a 1 to 10 scale divided into 10% increments with a 10 reflecting 91-100% defoliation.
$ LI-COR LI-3000 portable leaf area meter measurements in cm2 of leaf tissue.































APPENDIX B
LIST OF DEFOLIATION SCORES GENERATED BY WHOLE PLANT VISUAL,
PARTITIONED PLANT VISUAL, AND LEAF AREA OF THE PARTITIONED
PLANT RATING METHODS IN 1991.










List of defoliation scores generated by whole plant visual, partitioned plant visual,
of the partitioned plant rating methods.


and leaf area


Observation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39


Cage
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1


Plant
9
197
214
15
250
235
63
180
61
42
223
150
186
26
225
187
3
248
17
196
222
275
8
158
166
140
141
138
85
244
122
144
105
92
126
67
51
75
80


Whole plant
3
5
4
8
-
4
6
2
3
4
3
6
8
2
3
7
2
8
9
8
4
5
2
7
7
8
4
2
5
4
3
8
5
6
8
7
6
7
5


Avg.
visual
2.00
4.00
3.00
7.00

4.33
6.67
2.00
3.33
3.67
3.67
5.00
7.67
3.00
2.67
5.67
3.33
6.67
7.67
8.00
3.67
7.00
3.67
5.67
7.67
8.00
5.33
3.67
5.00
4.33
5.33
8.33
5.33
6.67
8.33
7.00
8.67
8.33
6.33


Avg.
LICOR
1.67
2.00
3.00
5.00

2.00
5.00
1.33
2.67
2.00
4.00
4.00
5.00
2.33
2.33
4.67
1.33
6.33
7.00
6.00
3.00
5.00
1.67
4.33
4.00
5.67
2.33
2.33
3.33
3.00
2.00
6.00
3.00
4.67
6.00
5.00
6.00
4.67
3.33





--------


Observation
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80


Cage
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2


Plant
178
179
22
110
270
212
100
213
25
103
79
248
213
17
270
150
100
194
186
61
79
9
250
26
222
214
244
103
178
22
85
180
42
110
212
275
158
51
75
25
8


Whole plant
4
5
8
7
9

6
3
2
5
3
4
2
8
8
5
4
6
6
7
4
4
9
2
7
1
2
5
3
8
7
3
3
5

5
6
4
5
5
3


Avg.
visual
5.33
6.00
7.67
6.33
8.67

5.67
4.33
4.00
5.67
3.67
3.33
3.33
8.33
9.00
5.67
5.00
6.67
5.67
7.67
4.00
3.00
9.33
2.33
6.00
3.33
4.00
6.33
4.67
8.67
5.33
5.00
6.33
4.67

5.00
7.00
4.33
6.33
2.00
3.00


Avg.
LICOR
3.33
4.00
4.67
5.00
7.00

2.67
2.00
3.00
2.67
3.00
1.67
3.33
5.67
4.33
4.00
3.00
3.33
4.00
5.00
2.67
4.33
6.33
1.67
5.00
2.00
2.00
3.00
3.00
5.33
5.33
2.67
5.00
3.33

3.33
3.00
3.00
3.33
1.67
1.67










Avg. Avg.
Observation Cage Plant Whole plant visual LICOR
81 2 140 8 4.67 2.33
82 2 15 3 8.67 5.00
83 2 138 3 2.67 1.33
84 2 67 6 6.67 3.33
85 2 126 7 8.00 6.00
86 2 63 8 7.67 5.33
87 2 179 6 7.33 5.33
88 2 80 7 7.67 2.67
89 2 92 6 6.67 2.00
90 2 3 3 4.33 2.00
91 2 223 4 6.67 3.33
92 2 166 5 7.00 4.00
93 2 122 2 3.67 2.33
94 2 225 3 4.67 2.00
95 2 196 9 9.00 5.33
96 2 144 9 9.00 5.00
97 2 187 3 5.33 2.33
98 2 105 4 6.33 3.00
99 2 141 3 4.33 3.00
100 2 235 5 7.67 3.00
Mean (N=97) 5.10 5.65 3.62
Variance 4.70 3.88 2.19
Std.Dev. 2.17 1.97 1.48
t Based on a 1 to 10 scale divided into 10% increments with a 10 reflecting 91-100% defoliation.
$ Based on LI-COR LI-3000 portable leaf area meter measurements that have been converted to %
defoliation and categorized into the 1-to-10 scale.
Missing data points.













APPENDIX C
GENERATION MEANS ANALYSIS
SAS PROGRAM


OPTIONS LS=80 PS =60;
/*This is a Generation Means Analysis based on Hayman's methodology, Heredity
12:371-390 */
DATA;
INFILE 'A:Cagel.dat';
/* The INPUT statement will vary according to the data set, you need generation
("GEN") and a dependent variable */
INPUT row plant gen $ fc$ rating l rating rating;
MEANRATE = (rating 1+rating2+rating3)/3;
PROC SORT; BY GEN;
PROC MEANS; BY GEN; VAR MEANRATE;
OUTPUT OUT=NE MEAN=Y VAR=V STDERR=S;
DATA NEW; SET NE; RS=1/S; IF GEN='D' OR GEN='BD' THEN DELETE;
PROC PRINT;
/* You need a minimum of 6 generations to conduct this analysis, if the number of
generations used are different than the amount (9) in this example, refer to Gamble's
paper (Canadian J. Plant Sci. 42:339-348) to obtain the proper coefficients. Another
source of information is Jennings, et al. (IA State J. Research, 48:267-280) */
DATA COEFCNTS;

INPUT GEN $ X1 X2 X3 X4 X5;
CARDS;
BC1 0.5 0.0 0.25 0.0 0.0
BC2 -0.5 0.0 0.25 0.0 0.0
BS1 0.5 -0.25 0.25 -0.25 0.0625
BS2 -0.5 -0.25 0.25 0.25 0.0625
F1 0.0 0.5 0.0 0.0 0.25
F2 0.0 0.0 0.0 0.0 0.0
F3 0.0 -0.25 0.0 0.0 0.0625
P1 1.0-0.5 1.0 -1.0 0.25
P2 -1.0-0.5 1.0 1.0 0.25

DATA FINAL; MERGE NEW COEFCNTS;
PROC PRINT;









/* This model tests the significance of the additive (X1) and the dominant (X2) effects
*/
TITLE 'PROC REG WEIGHTED';
PROC REG;
MODEL Y = X1 X2;
WEIGHT RS;
/* The model must be weighted to account for the unequal population sizes among the
generations. See Rowe and Alexander, Crop Sci. 20:109-110, for further detail. The next
model tests for the epistatic effects, additive-additive (X3), additive-dominant (X4), and
dominant-dominant (X5), as well as the additive (X1) and dominant (X2) effects. */
PRO REG;
MODEL Y = X1 X2 X3 X4 X5;
WEIGHT RS;
RUN;

/* This next set of models are tested with PROC GLM instead of PROC REG */
TITLE 'PROC GLM WEIGHTED';
PROC GLM;
MODEL Y = X1 X2;
WEIGHT RS;
PROC GLM;
MODEL Y = X1 X2 X3 X4 X5;
WEIGHT RS;
RUN;

/* The next series of models are the same as the previous models, except that they
assume equal population size among the generations, and are therefore not weighted. */
TITLE 'PROC REG NOWEIGHT';
PROC REG;
MODEL Y = X1 X2;
PROC REG;
MODEL Y = X1 X2 X3 X4 X5;
RUN;
TITLE 'PROC GLM NOWEIGHT';
MODEL Y = X1 X2;
PROC GLM;
MODEL Y = Xl X2 X3 X4 X5;
RUN;
/* Depending on whether or not the population sizes among the generations are equal,
report the model that has the most significant effects from either PROC GLM or PROC
REG. */




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EOKF01386_UCGUX0 INGEST_TIME 2011-09-29T20:14:03Z PACKAGE AA00004723_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

,1+(5,7$1&( 2) 5(6,67$1&( 72 7+( 62<%($1 /223(5 ,1 62<%($1 %\ 0,&+$(/ 0217*20(5< .(17< $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

, ZRXOG OLNH WR GHGLFDWH WKLV ZRUN WR WKH PHPRU\ RI WKH ODWH 'U -RKQQ\ 2X]WV 7KDQN \RX IRU \RXU IULHQGVKLS DQG JXLGDQFH ZLWKRXW LW ZRXOG QRW KDYH PDGH LW WKLV IDU

PAGE 3

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

PAGE 4

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

PAGE 5

$33(1',; $ /,67 2) '()2/,$7,21 6&25(6 *(1(5$7(' %< :+2/( 3/$17 9,68$/ 3$57,7,21(' 3/$17 9,68$/ $1' /($) $5($ 2) 7+( 3$57,7,21(' 3/$17 5$7,1* 0(7+2'6 ,1 $33(1',; % /,67 2) '()2/,$7,21 6&25(6 *(1(5$7(' %< :+2/( 3/$17 9,68$/ 3$57,7,21(' 3/$17 9,68$/ $1' /($) $5($ 2) 7+( 3$57,7,21(' 3/$17 5$7,1* 0(7+2'6 ,1 $33(1',; & *(1(5$7,21 0($16 $1$/<6,6 6$6 352*5$0 $33(1',; *(1(5$7,21 0($16 $1$/<6,6 6$6 352*5$0 $33(1',; ( 7(67 )25 1250$/,7< 5()(5(1&(6 %,2*5$3+,&$/ 6.(7&+ Y

PAGE 6

/,67 2) 7$%/(6 SDJH 7DEOH 0F1HPDU WHVW IRU VLJQLILFDQW GLIIHUHQFHV EHWZHHQ WKH :3 ZKROH SODQW YLVXDOf DQG $9 SDUWLWLRQHG SODQW YLVXDOf PHWKRGV LQ DQG EHWZHHQ WKH :3 DQG $9 :3 DQG $/ OHDI DUHD RI WKH SDUWLWLRQHG SODQWf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fV VFDOLQJ WHVW DSSOLHG WR WKH FURVV [ %UD[WRQ WR WHVW DGHTXDF\ RI DGGLWLYHGRPLQDQFH PRGHO IRU UHVLVWDQFH WR VR\EHDQ ORRSHU YL

PAGE 7

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

PAGE 8

/,67 2) ),*85(6 (JH )LJXUH ([DPSOHV RI SKRWRFRSLHG OHDIOHWV XVHG LQ GHWHUPLQLQJ WKH b GHIROLDWLRQ LQ WKH $9 OHDI DUHD RI SDUWLWLRQHG SODQWf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

PAGE 9

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ ,1+(5,7$1&( 2) 5(6,67$1&( 72 7+( 62<%($1 /223(5 ,1 62<%($1 %\ 0LFKDHO 0RQWJRPHU\ .HQW\ $SULO &KDLUPDQ .XHOO +LQVRQ 0DMRU 'HSDUWPHQW $JURQRP\ 3K\WRSKDJRXV LQVHFWV FDXVH PLOOLRQV RI GROODUV RI GDPDJH WR VR\EHDQ >*O\FLQH PD[ /f 0HUU@ WKURXJKRXW WKH VRXWKHUQ 8QLWHG 6WDWHV 7KLV VWXG\ ZDV FRQGXFWHG WR f HYDOXDWH WKUHH PHWKRGV IRU UDWLQJ GHIROLDWLRQ RI VR\EHDQ E\ LQVHFWV DQG LLf GHWHUPLQH WKH LQKHULWDQFH RI UHVLVWDQFH WR WKH VR\EHDQ ORRSHU >3VHXGRSOXVLD LQFOXGHQV :DONHU@ LQ VR\EHDQ 7KH LQVHFWUHVLVWDQW EUHHGLQJ OLQH ZDV FURVVHG ZLWK WKH FXOWLYDU %UD[WRQ LQ $GGLWLRQDO FURVVHV ZHUH PDGH LQ DQG WR SURYLGH WKH SRSXODWLRQV QHFHVVDU\ WR FRQGXFW UDWLQJ PHWKRGRORJ\ DQG JHQHWLF VWXGLHV LQ ILHOG FDJHV LQ DQG 7KH WKUHH UDWLQJ PHWKRGV HYDOXDWHG ZHUH ZKROH SODQW YLVXDO :3f SDUWLWLRQHG SODQW DYHUDJH YLVXDO $9f DQG PHDVXUHG DYHUDJH OHDI DUHD RI SDUWLWLRQHG SODQW $/f ,Q WKH :3 DQG $/ ZHUH KLJKO\ FRUUHODWHG U S f DQG LQ DOO FRUUHODWLRQV :3 YV $9 :3 YV $/ DQG $9 YV $/f ZHUH VLJQLILFDQW DW S ,;

PAGE 10

%DVHG RQ WKH UHODWLYH YDULDWLRQ RI HDFK PHWKRG WKH SUHFLVLRQ RI WKH WKUHH PHWKRGV ZDV HVVHQWLDOO\ HTXDO DQG YHU\ JRRG ZHOO EHORZ WKH GHVLUDEOH OHYHO RI 7KHVH UHVXOWV LQGLFDWHG WKDW UHODWLYH HVWLPDWHV DUH VXLWDEOH IRU JHQHWLF VWXGLHV ,Q WKH LQKHULWDQFH VWXG\ GHIROLDWLRQ ZDV HVWLPDWHG E\ WKH :3 PHWKRG LQ DQG WKH $9 PHWKRG LQ 7KH GDWD IURP WKH SUHOLPLQDU\ VWXG\ LQ VKRZHG D WUHQG WRZDUGV TXDQWLWDWLYH LQKHULWDQFH WKHUHIRUH WKH GDWD ZHUH DQDO\]HG TXDQWLWDWLYHO\ 0DWKHUfV VFDOLQJ WHVW ZDV DSSOLHG WR WKH GDWD JHQHUDWHG IURP HDFK FDJH DQG WKH UHVXOWV LQGLFDWHG WKDW JHQHUDWLRQ PHDQV GHSHQG RQO\ RQ DGGLWLYH JHQH HIIHFWV 8WLOL]LQJ +D\PDQfV PHWKRGRORJ\ LQ WHUPV RI WKH JHQHUDWLRQ PHDQV DQDO\VLV DQ HSLVWDWLF HIIHFW ZDV VXJJHVWHG EXW WKH SULPDU\ HIIHFW LV DVVXPHG WR EH DGGLWLYH DV ZDV LQGLFDWHG LQ WKH VFDOLQJ WHVW (VWLPDWHV RI JHQH QXPEHUV LQGLFDWH WKDW WKH WZR SDUHQWV GLIIHUHG E\ WZR JHQHV IRU UHVLVWDQFH WR VR\EHDQ ORRSHU +HULWDELOLW\ IRU UHVLVWDQFH ZDV HVWLPDWHG WR EH b WKHUHIRUH D EUHHGHU VKRXOG EH DEOH WR PDNH SURJUHVV E\ VHOHFWLQJ LQ WKH ) RU ) SRSXODWLRQV 7KH SRWHQWLDO H[LVWV IRU LQFUHDVLQJ UHVLVWDQFH WR WKH VR\EHDQ ORRSHU DQG SRVVLEO\ RWKHU SK\WRSKDJRXV LQVHFWV E\ S\UDPLGLQJ JHQHV IURP QHZ VRXUFHV RI UHVLVWDQFH [

PAGE 11

&+$37(5 ,1752'8&7,21 6R\EHDQ >*O\FLQH PD[ /f 0HUULOO@ LV D PDMRU ZRUOG FURS 7RWDO ZRUOG KHFWDUHV SODQWHG LQFUHDVHG IURP PLOOLRQ LQ WR PLOOLRQ LQ f 7KLV b LQFUHDVH ZDV GXH SULPDULO\ WR GHPDQG IRU RLO DQG PHDO SURGXFWV DQG WR D VPDOO H[WHQW WKH ZKROH EHDQ SURGXFW f ,Q WKH 8QLWHG 6WDWHV KHFWDUDJH URVH WR D SHDN RI PLOOLRQ KHFWDUHV LQ ZLWK D VWHDG\ GHFOLQH WR PLOOLRQ KHFWDUHV LQ 7KLV b GHFUHDVH LQ VR\EHDQ KHFWDUDJH ZDV SULPDULO\ GXH WR UHGXFHG SODQWLQJV LQ WKH VRXWKHDVWHUQ 8QLWHG 6WDWHV f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f 6LPLODU ORVVHV DOVR RFFXUUHG LQ VWDWHV QHLJKERULQJ 0LVVLVVLSSL DV D UHVXOW RI LQVHFW LQIHVWDWLRQV f 3HVW SUREOHPV ZKLFK FDQ EH FRVWO\ WR FRQWURO DQG D \HDU f DYHUDJH SULFH RI SHU EX f PDNH LW LQFUHDVLQJO\ GLIILFXOW IRU D JURZHU WR SURGXFH D SURILWDEOH FURS

PAGE 12

6HYHUDO WHFKQLTXHV DUH DYDLODEOH IRU FRQWURO RI GLVHDVHV QHPDWRGHV DQG LQVHFWV LQ VR\EHDQ 0HWFDOI DQG /XFNPDQQ f FDWHJRUL]HG WKHVH WHFKQLTXHV LQWR VHYHQ PDMRU PHWKRGV FXOWXUDO PHFKDQLFDO SK\VLFDO ELRORJLFDO FKHPLFDO JHQHWLF DQG UHJXODWRU\ $OO PHWKRGV RI FRQWURO KDYH PHULW DQG VKRXOG EH XWLOL]HG LQ LQWHJUDWHG SHVW PDQDJHPHQW SURJUDPV f $OWKRXJK HFRQRPLFV LV WKH PDMRU OLPLWLQJ IDFWRU LQ DFKLHYLQJ FRQWURO RI DQ\ SHVW RI VR\EHDQ KXPDQ VDIHW\ DQG HQYLURQPHQWDO LPSDFW PXVW EH FRQVLGHUHG ZKHQ FKRRVLQJ DQ DSSURDFK WR FRQWURO D SDUWLFXODU SHVW %HIRUH PDNLQJ D GHFLVLRQ RQ ZKDW DSSURDFK WR WDNH D JURZHU VKRXOG DQVZHU WKH TXHVWLRQV f:LOO WKH HQG UHVXOWV MXVWLI\ WKH FRVW RI FRQWURO"f DQG f,V WKH FRQWURO PHDVXUH VDIH"f 7KH VR\EHDQ ORRSHU KDV EHFRPH LQFUHDVLQJO\ UHVLVWDQW WR DYDLODEOH LQVHFWLFLGHV f )RU FRQWURO RI WKH VR\EHDQ ORRSHU DQG RWKHU LQVHFW SHVWV D ELRORJLFDO DSSURDFK VKRXOG \LHOG WKH EHVW UHVXOWV 7KHUH DUH WZR ELRORJLFDO DSSURDFKHV WR LQVHFW FRQWURO KRVW SODQW UHVLVWDQFH DQG WKH XVH RI ELRFRQWURO DJHQWV f 7KH ODWWHU LQYROYHV WKH LGHQWLILFDWLRQ DQG H[SORLWDWLRQ RI QDWXUDO HQHPLHV DQG SDWKRJHQV 7KLV PHWKRG RI FRQWURO XVXDOO\ LV OLPLWHG WR D VSHFLILF LQVHFW SHVW UDWKHU WKDQ WKH EURDG VSHFWUXP RI LQVHFWV WKDW GDPDJH VR\EHDQ $V ZLWK FRQYHQWLRQDO LQVHFWLFLGHV UHVLVWDQFH WR PLFURELDO LQVHFWLFLGHV KDV EHHQ REVHUYHG f +RVW SODQW UHVLVWDQFH RIIHUV DQ HIIHFWLYH PHWKRG RI FRQWURO ZKLFK LV EXLOW LQWR WKH SODQW ,W KDV UHODWLYHO\ ORQJ VWDELOLW\ LV FRPSDWLEOH ZLWK RWKHU FRQWURO WDFWLFV LV HQYLURQPHQWDOO\ VDIH f DQG GHYHORSPHQW DQG VXEVHTXHQW UHOHDVH RI D UHVLVWDQW FXOWLYDU LV IDU PRUH VRFLDOO\ DFFHSWDEOH f /DPEHUW DQG .LOHQ f IRXQG WKDW GLUHFW VHOHFWLRQ IRU UHVLVWDQFH LQ VR\EHDQ WR RQH VSHFLHV RI LQVHFW UHVXOWHG LQ WKH LQGLUHFW VHOHFWLRQ IRU

PAGE 13

UHVLVWDQFH WR FHUWDLQ RWKHU VSHFLHV 7KH GHYHORSPHQW RI DQ LQVHFWUHVLVWDQW VR\EHDQ FXOWLYDU WDNHV DSSUR[LPDWHO\ WKH VDPH DPRXQW RI WLPH DV LV UHTXLUHG WR GHYHORS D QHZ HIIHFWLYH LQVHFWLFLGH ZKLFK FDQ EH DV VKRUW DV WR \HDUV RU DV ORQJ DV \HDUV 7KH LQVHFWUHVLVWDQW FXOWLYDUV f/DPDUf f DQG f&URFNHWWf f HDFK WRRN DSSUR[LPDWHO\ \HDUV WR GHYHORS ,QLWLDOO\ 9DQ 'X\Q HW DO f LGHQWLILHG WKH WKUHH SODQW LQWURGXFWLRQV 3, 3, DQG 3, IURP WKH 86'$ JHUPSODVP FROOHFWLRQ DV KDYLQJ WKH KLJKHVW OHYHO RI UHVLVWDQFH WR WKH 0H[LFDQ EHDQ EHHWOH ^(SLODFKQD YDULYHVWLV 0XOVDQWf 6FKLOOLQJHU f LGHQWLILHG DGGLWLRQDO VRXUFHV RI UHVLVWDQFH WR VSHFLILF LQVHFWV DQG GHVFULEHG WKH PHFKDQLVP RI UHVLVWDQFH WKDW HDFK H[KLELWV +DWFKHWW HW DO f GHWHUPLQHG WKDW EUHHGLQJ OLQHV LGHQWLILHG DV SRVVHVVLQJ UHVLVWDQFH WR PXOWLSOH VSHFLHV RI LQVHFWV FRXOG VHUYH DV EDVLF JHUPSODVP IRU GHYHORSPHQW RI FXOWLYDUV UHVLVWDQW WR LQVHFWV 7R GDWH WZR VR\EHDQ FXOWLYDUV DQG WZR VR\EHDQ JHUPSODVP OLQHV ZLWK UHVLVWDQFH WR LQVHFWV KDYH EHHQ UHJLVWHUHG ZLWK WKH &URS 6FLHQFH 6RFLHW\ RI $PHULFD f $OWKRXJK WKHUH KDV EHHQ VXFFHVV LQ GHYHORSLQJ LQVHFWUHVLVWDQW JHUPSODVP OLQHV DQG FXOWLYDUV WKHUH KDV EHHQ YHU\ OLWWOH UHVHDUFK FRQGXFWHG RQ WKH LQKHULWDQFH RI LQVHFW UHVLVWDQFH LQ VR\EHDQ f 2I WKH WKUHH LQKHULWDQFH VWXGLHV WZR f ZHUH FRQGXFWHG RQ WKH UHVLVWDQFH WR 0H[LFDQ EHDQ EHHWOH DQG RQH f RQ WKH UHVLVWDQFH WR WKH YHOYHWEHDQ FDWHUSLOODU $QWLFDUVLD JHPPDWDOLV +XEQHUf 6LQFH WKH VR\EHDQ ORRSHU LV D PDMRU SHVW WR VR\EHDQ LQ WKH VRXWKHDVWHUQ 8QLWHG 6WDWHV DQG LV GLIILFXOW WR FRQWURO ZLWK LQVHFWLFLGHV WKH LQKHULWDQFH RI UHVLVWDQFH WR WKH VR\EHDQ ORRSHU ZRXOG DLG EUHHGHUV LQ WKH GHYHORSPHQW RI UHVLVWDQW FXOWLYDUV

PAGE 14

7KH REMHFWLYHV RI WKLV VWXG\ ZHUH f WR HYDOXDWH WKUHH PHWKRGV IRU UDWLQJ GHIROLDWLRQ RI VR\EHDQ E\ VR\EHDQ ORRSHU DQG nnf WR GHWHUPLQH WKH LQKHULWDQFH RI UHVLVWDQFH LQ VR\EHDQ WR WKH VR\EHDQ ORRSHU

PAGE 15

&+$37(5 /,7(5$785( 5(9,(: $OWKRXJK XQGRFXPHQWHG SODQWV WKDW DUH UHVLVWDQW WR LQVHFWV KDYH VXUYLYHG DQG SURSDJDWHG EDVHG RQ WKH SURFHVVHV RI DGDSWDWLRQ DQG QDWXUDO VHOHFWLRQ 7KHVH fVXUYLYRUVf ZHUH LQ WXUQ VHOHFWHG E\ IDUPHUV WR SODQW IXWXUH FURSV f 7KLV SUDFWLFH RI VHOHFWLRQ E\ SULPLWLYH IDUPHUV FRXOG EH FRQVLGHUHG WKH EHJLQQLQJ RI WKH GHYHORSPHQW RI LQVHFW UHVLVWDQW FXOWLYDUV $FFRUGLQJ WR .RJDQ f 3HGLJR f DQG 6PLWK f LQ WKHLU EULHI KLVWRULHV RI LQVHFWUHVLVWDQW SODQWV WKH ILUVW GHYHORSPHQW DQG VXEVHTXHQW FXOWLYDWLRQ RI LQVHFWUHVLVWDQW FXOWLYDUV RFFXUUHG LQ WKH ODWH HLJKWHHQWK DQG HDUO\ QLQHWHHQWK FHQWXULHV ,Q -1 +DYHQV LGHQWLILHG WKH ZKHDW 7ULWLFXP DHVWLYLXP / HP 7KHOOf FXOWLYDU f8QGHUKLOOf DV EHLQJ UHVLVWDQW WR WKH +HVVLDQ IO\ 0D\HWLROD GHVWUXFWRU 6D\f ,Q WKH fV /LQGOH\ UHFRPPHQGHG WKH FXOWLYDWLRQ RI DSSOH 0DOXV SXPLOD 0LOOHUf FXOWLYDUV UHVLVWDQW WR WKH ZRROO\ DSSOH DSKLG (ULRVRPD ODQLJHUXP +DXVPDQf ,Q WKH PLGQLQHWHHQWK FHQWXU\ WKH )UHQFK ZLQH LQGXVWU\ ZDV VDYHG E\ JUDIWLQJ WKH KLJKO\ VXVFHSWLEOH VFLRQV RI WKH )UHQFK YDULHWLHV WR WKH URRWVWRFNV RI $PHULFDQ JUDSHV 9LWLV / VSSf ZKLFK ZHUH UHVLVWDQW WR WKH JUDSH SK\OOR[HUD 3K\OOR[HUD YLWWLIRODH )LWFKf ,Q WKH 8QLWHG 6WDWHV 5+ 3DLQWHU LV FRQVLGHUHG WR EH WKH IRXQGHU DQG SLRQHHU RI WKH PRGHP HUD RI UHVHDUFK RQ LQVHFWUHVLVWDQW SODQWV f ,Q KLV ERRN ,QVHFW

PAGE 16

5HVLVWDQFH LQ &URS 3ODQWV ZKLFK ZDV WKH ILUVW ERRN ZULWWHQ RQ WKH WRSLF KH GHVFULEHG PHFKDQLVPV RI UHVLVWDQFH LQ SODQWV DQG IDFWRUV WKDW DIIHFW WKHP %DVHG RQ ILHOG REVHUYDWLRQV KH VHSDUDWHG WKH PHFKDQLVPV RI UHVLVWDQFH LQ SODQWV LQWR WKUHH GLVWLQFW EXW LQWHUDFWLYH FODVVHV QRQSUHIHUHQFH ODWHU NQRZQ DV DQWL[HQRVLV f IRU RYLSRVLWLRQ IRRG RU VKHOWHUf DQWLELRVLV DGYHUVH HIIHFW RQ WKH ELRORJ\ RI WKH LQVHFWf DQG WROHUDQFH DELOLW\ WR ZLWKVWDQG UHSDLU RU UHFRYHU IURP LQIHVWDWLRQf f 'HYHORSLQJ D FXOWLYDU WKDW H[KLELWV D KLJK OHYHO RI DOO WKUHH PHFKDQLVPV RI UHVLVWDQFH ZRXOG EH RSWLPXP 7KLV LV VHOGRP IHDVLEOH WKHUHIRUH GHYHORSLQJ FXOWLYDUV ZLWK WKH DQWLELRVLV PHFKDQLVP RI UHVLVWDQFH LV WKH PRVW IUHTXHQW REMHFWLYH RI SODQW EUHHGHUV $OWKRXJK QRQSUHIHUHQFH DQG DQWLELRVLV GLIIHU LQ WKHLU PHFKDQLVPV RI UHVLVWDQFH ERWK LQYROYH WKH LQWHUDFWLRQ RI WKH ELRORJ\ RI WKH SODQW ZLWK WKH ELRORJ\ RI WKH LQVHFW f 7KHUH LV RIWHQ DQ RYHUODSSLQJ RI WKH QRQSUHIHUHQFH DQG DQWLELRVLV FODVVHV ZKLFK PDNHV LW GLIILFXOW WR GHWHUPLQH WKH H[DFW PHFKDQLVP RI UHVLVWDQFH H[KLELWHG LQ D SODQW WR D SDUWLFXODU VSHFLHV RI LQVHFW f $OWKRXJK LW LV DGYDQWDJHRXV LI D EUHHGHU NQRZV ZKLFK PHFKDQLVP RI UHVLVWDQFH LV EHLQJ H[SUHVVHG NQRZOHGJH RI WKH PHFKDQLVP LV QRW HVVHQWLDO WR PDNH DGYDQFHV LQ WKH GHYHORSPHQW RI UHVLVWDQW JHUPSODVP ,Q IDFW UHVLVWDQW FXOWLYDUV DQG JHUPSODVP KDYH EHHQ GHYHORSHG IRU \HDUV ZLWK OLWWOH FRQFHUQ IRU WKH XQGHUO\LQJ PHFKDQLVP RI UHVLVWDQFH ,W KDV RQO\ EHHQ LQ WKH SDVW WZR GHFDGHV WKDW PHFKDQLVPV RI UHVLVWDQFH KDYH EHHQ LGHQWLILHG LQ FURS SODQWV f ,Q VR\EHDQ 9DQ 'X\Q HW DO f LGHQWLILHG WKUHH JHQRW\SHV 3, 3, DQG 3, f IURP WKH 86'$ JHUPSODVP FROOHFWLRQ RI PDWXULW\ JURXSV 9,, DQG 9,,, DV UHVLVWDQW WR WKH 0H[LFDQ EHDQ EHHWOH 7KHVH WKUHH 3LfV KDYH EHHQ XVHG

PAGE 17

H[WHQVLYHO\ DV VRXUFHV RI UHVLVWDQW JHUPSODVP DQG DV VWDQGDUG FKHFNV LQ VFUHHQLQJ DGGLWLRQDO JHUPSODVP IRU UHVLVWDQFH $W WKH :RUOG 6R\EHDQ 5HVHDUFK &RQIHUHQFH LQ 6FKLOOLQJHU f UHSRUWHG DGGLWLRQDO VRXUFHV RI UHVLVWDQFH WR WKH 0H[LFDQ EHDQ EHHWOH *DU\ HW DO f VFUHHQHG OLQHV IURP PDWXULW\ JURXSV 9, 9,, 9,,, DQG ,; RI WKH 86'$ JHUPSODVP FROOHFWLRQ DJDLQVW YHOYHWEHDQ FDWHUSLOODU LQLWLDOO\ DQG VXEVHTXHQWO\ YHOYHWEHDQ FDWHUSLOODU VR\EHDQ ORRSHU FRUQ HDUZRUP +HOLRWKLV ]HD %RGGLHf WREDFFR EXGZRUP + YLUHVFHQV )DEULFLXVf DQG EHHW DUP\ ZRUP 6SRGRSWHUD H[LJXD +XEQHUf %DVHG RQ WKH UHVXOWV RI WKH VFUHHQLQJV DJDLQVW WKHVH ILYH VSHFLHV 3, DQG )& ZHUH LGHQWLILHG DV KDYLQJ OHYHOV RI UHVLVWDQFH WKDW ZRXOG EH VXLWDEOH IRU GHYHORSLQJ UHVLVWDQW FXOWLYDUV 7KH JHUPSODVP FROOHFWLRQ RI 3LfV RI ZLOG VR\EHDQ *O\FLQH VRMD 6LHE t =XFF ZDV VFUHHQHG DJDLQVW VR\EHDQ ORRSHU YHOYHWEHDQ FDWHUSLOODU EHHW DUP\ ZRUP DQG FRP HDUZRUP WR LGHQWLI\ DGGLWLRQDO VRXUFHV RI UHVLVWDQFH HTXDO WR RU JUHDWHU WKDQ 3LfV DQG 0F.HQQD HW DO f LGHQWLILHG WKUHH OLQHV 3LfV DQG f IURP WKLV VWXG\ DV KDYLQJ OHVV GHIROLDWLRQ WKDQ WKH WZR VWDQGDUGV WKXV VXJJHVWLQJ WKDW WKHUH DUH DGGLWLRQDO JHQHV IRU LQVHFW UHVLVWDQFH ,Q .UDHPHU HW DO f LGHQWLILHG VHYHUDO DFFHVVLRQV RI WKH 86'$ JHUPSODVP FROOHFWLRQ DV EHLQJ UHVLVWDQW WR WKH 0H[LFDQ EHDQ EHHWOH 7KH\ ZHUH IURP PDWXULW\ JURXSV 9, )& DQG 3LfV DQG f DQG 9,,, 3LfV DQG f 8SRQ LGHQWLILFDWLRQ RI D JHQRW\SH WKDW LV UHVLVWDQW WR D FHUWDLQ LQVHFW VSHFLHV WKH JHQRW\SH XVXDOO\ LV VFUHHQHG DJDLQVW RWKHU LQVHFW VSHFLHV 5HVLVWDQFH WR PRUH WKDQ RQH VSHFLHV ZLOO HQKDQFH LWV XVHIXOQHVV LQ D EUHHGLQJ SURJUDP +DWFKHWW HW DO f FRQGXFWHG

PAGE 18

D VWXG\ XWLOL]LQJ EUHHGLQJ OLQHV WKDW GHULYH WKHLU LQVHFW UHVLVWDQFH IURP 3, WR GHWHUPLQH LI VHOHFWLRQ IRU UHVLVWDQFH WR RQH RU WZR LQVHFW VSHFLHV ZRXOG UHVXOW LQ WKH LQGLUHFW VHOHFWLRQ IRU UHVLVWDQFH WR RWKHU WURXEOHVRPH VSHFLHV 7KH\ IRXQG WKDW QR VLQJOH LQVHFW VSHFLHV FRXOG EH XWLOL]HG WR VHOHFW IRU UHVLVWDQFH WR PXOWLSOH VSHFLHV RI LQVHFWV DPRQJ DGDSWHG JHQRW\SHV ,Q D VLPLODU VWXG\ /DPEHUW DQG .LOHQ f GHWHUPLQHG WKDW ZKHQ 3, ZDV XVHG DV WKH GRQRU SDUHQW GLUHFW VHOHFWLRQ IRU UHVLVWDQFH WR RQH LQVHFW VSHFLHV UHVXOWHG LQ LQGLUHFW VHOHFWLRQ IRU UHVLVWDQFH WR RWKHU IROLDUIHHGLQJ VSHFLHV 5HVHDUFK WR GHYHORS VR\EHDQ JHQRW\SHV UHVLVWDQW WR LQVHFWV LV XVXDOO\ FRQGXFWHG ZLWK VSHFLHV WKDW DUH LQGLJHQRXV WR D SDUWLFXODU DUHD f /DPEHUW DQG .LOHQ f HYDOXDWHG 3, 3, DQG 3, DJDLQVW ILYH LQVHFW VSHFLHV WR GHWHUPLQH WKHLU UHODWLYH OHYHOV RI UHVLVWDQFH ,QVHFW VSHFLHV WHVWHG ZHUH YHOYHWEHDQ FDWHUSLOODU VR\EHDQ ORRSHU FRP HDUZRUP WREDFFR EXG ZRUP DQG EHHW DUP\ ZRUP %DVHG XSRQ WKH UHVSRQVHV WR WKHVH ILYH VSHFLHV WKH\ IRXQG WKDW WKH PHFKDQLVP IRU UHVLVWDQFH LQ 3, ZDV QRQSUHIHUHQFH ZKHUHDV WKDW RI 3, RU 3, ZDV DQWLELRVLV 7KHVH VDPH WKUHH 3LfV ZHUH HYDOXDWHG DJDLQVW IRXU LPSRUWDQW VR\EHDQ GHIROLDWLQJ VSHFLHV EHHW DUP\ ZRUP 3RUWKHVLD WDLZDQLD 6KLUDNL $QPDOD FXSULSHV %DWHV DQG 2UJ\LD VSHFLHVf WR GHWHUPLQH WKHLU XVHIXOQHVV LQ GHYHORSLQJ EUHHGLQJ OLQHV UHVLVWDQW WR LQVHFWV LQGLJHQRXV WR 7DLZDQ 1R VLQJOH DFFHVVLRQV KDG D FRQVLVWHQWO\ KLJK OHYHO RI UHVLVWDQFH WR DOO RI WKH LQVHFW VSHFLHV 3, GHPRQVWUDWHG WKH KLJKHVW OHYHO RI UHVLVWDQFH WR WKH PRVW SUHYDOHQW VSHFLHV WKH EHHW DUP\ZRUP DQG WKHUHIRUH ZRXOG EH VXLWDEOH WR LQLWLDWH D EUHHGLQJ SURJUDP f

PAGE 19

:LWK WKH VRXUFHV RI UHVLVWDQFH WR YDULRXV VSHFLHV RI LQVHFWV LGHQWLILHG UHVHDUFKHUV KDYH FKDQQHOOHG WKHLU HQHUJ\ LQWR GHWHUPLQLQJ WKH EDVLV RI WKH PHFKDQLVP RI UHVLVWDQFH LQ FHUWDLQ VR\EHDQ DFFHVVLRQV 6LQFH WKH PLGfV UHVHDUFKHUV KDYH LQYHVWLJDWHG YDULRXV FRQVWLWXHQWV RI VR\EHDQ WR LGHQWLI\ WKH XQGHUO\LQJ ELRFKHPLFDO FRPSRXQG WKDW FRQIHUV UHVLVWDQFH WR LQVHFWV f ,I DQ DVVRFLDWLRQ FRXOG EH HVWDEOLVKHG EHWZHHQ D SDUWLFXODU FRPSRXQG RU FODVV RI FRPSRXQGV DQG UHVLVWDQFH WKHQ D QHZ WRRO FRXOG EH GHYHORSHG WR DFFHOHUDWH WKH VHOHFWLRQ SURFHVV E\ EUHHGHUV ,Q VWXGLHV ZLWK 3, 3, DQG WZR VXVFHSWLEOH FXOWLYDUV 7HVWHU f UHSRUWHG WKDW UHVLVWDQW 3LfV KDG ORZHU WRWDO QLWURJHQ PRUH VROXEOH FDUERK\GUDWHV DQG PRUH WRWDO VWHUROV DW IORZHULQJ DQG SRGILOO WKDQ VXVFHSWLEOH FXOWLYDUV 7KURXJK D VHULHV RI JUDIWLQJ H[SHULPHQWV /DPEHUW DQG .LOHQ f YHULILHG WKDW FRQVWLWXHQWV FRQWULEXWLQJ WR UHVLVWDQFH LQ 3, DQG 3, ZHUH IRUPHG LQ WKH OHDYHV DQG WUDQVORFDWHG WKURXJKRXW WKH SODQW 6PLWK f GHWHUPLQHG WKDW WKH UHVLVWDQW IDFWRUV LQ 3, DSSHDUHG WR EH FKHPLFDO +H DOVR IRXQG WKDW UHVLVWDQFH WR WKH VR\EHDQ ORRSHU LQ 3, ZDV DQ DQWLELRWLF HIIHFW GXH WR WKH FRPELQHG HIIHFWV RI D IHHGLQJ GHWHUUHQW DQG D JURZWK LQKLELWRU &KLDQJ HW DO f VXJJHVWHG WKDW UHVLVWDQFH WR 0H[LFDQ EHDQ EHHWOH IHHGLQJ LQ 3, PD\ EH DWWULEXWHG WR SKHQ\OSURSDQRLG PHWDEROLWHV ,Q WKHLU LQRUJDQLF QXWULHQW DQDO\VLV RI OHDI WLVVXH VWXGLHV ZLWK VR\EHDQ UHVLVWDQW WR WKH 0H[LFDQ EHDQ EHHWOH 0HEUDKWX HW DO f IRXQG WKDW WKH FRQWHQW RI FDOFLXP SKRVSKRUXV DQG SRWDVVLXP ZDV QHJDWLYHO\ FRUUHODWHG ZLWK SXSDO ZHLJKW JDLQ U rr U rr DQG U r S DQG S UHVSHFWLYHO\f

PAGE 20

,Q .RJDQ f SXEOLVKHG D WUHDWLVH RQ WKH UROH RI ELRFKHPLFDOV LQ SODQW UHVLVWDQFH WR LQVHFWV +H H[DPLQHG WKH GHIHQVLYH SDWWHUQV RI WKUHH IDPLOLHV 6RODQDFHDH &UXFLIHUDH DQG /HJXPLQRVDH VSHFLILFDOO\ WKH FRPPRQ EHDQ DQG VR\EHDQf DQG SXEOLVKHG DV TXRWHG EHORZ D VXPPDWLRQ RI DOO SUHYLRXV UHVHDUFK ILQGLQJV WKDW HOXFLGDWHG WKH FKHPLFDO EDVHV RI UHVLVWDQFH LQ 3, 3, DQG 3, )ROLDU WRWDO QLWURJHQ VROXEOH FDUERK\GUDWHV RUJDQLF DFLGV DQG VWHUROV RI ERWK UHVLVWDQW DQG VXVFHSWLEOH VR\EHDQ YDU\ ZLWK VWDJH RI JURZWK 7HVWHU *UXQZDOG DQG .RJDQ f 7RWDO QLWURJHQ ZDV JHQHUDOO\ ORZHU DQG DW WKH HQG RI WKH VHDVRQ VROXEOH FDUERK\GUDWHV DQG VWHUROV ZHUH KLJKHU LQ WKH UHVLVWDQW OLQHV RUJDQLF DFLGV IOXFWXDWHG ZLWKRXW FRUUHODWLRQ ZLWK UHVLVWDQFH 7HVWRU f 6WHURO SURILOHV ZHUH VLPLODU LQ ERWK VXVFHSWLEOH DQG UHVLVWDQW OLQHV ERWK LQ HDUO\ DQG LQ ODWH PDWXULQJ VR\EHDQV *UXQZDOG DQG .RJDQ f 3LQLWRO LV D GRPLQDQW F\FOLWRO LQ VR\EHDQ DQG LW UHSUHVHQWV DQ DYHUDJH RI b RI WKH WRWDO bHWKDQROVROXEOH FDUERK\GUDWH IUDFWLRQ RI VR\EHDQ 3KLOOLSV HW DO f 3LQLWRO ZDV IRXQG LQ ODUJH DPRXQWV LQ 3, DQG LQ VPDOOHU FRQFHQWUDWLRQV LQ WKH VXVFHSWLEOH FXOWLYDU f'DYLVf # b GU\ ZHLJKWf $GGHG WR DUWLILFLDO PHGLD DW b SLQLWRO UHGXFHG ZHLJKW JDLQ E\ FRUQ HDUZRUP ODUYDH E\ DERXW b 7KHVH ILQGLQJV VXJJHVW WKDW SLQLWRO LV DQ DQWLELRWLF IDFWRU LQ VR\EHDQ EXW ZLWK QR DSSDUHQW DQWL[HQRWLF SURSHUW\ 'UH\HU HW DO f $V WKH UHSRUWHG UHVXOWV FRXOG QRW EH UHSOLFDWHG HOVHZKHUH WKLV FODLP KDV EHHQ UHFHQWO\ GLVSXWHG *DUGQHU HW DO f 6\VWHPDWLF IUDFWLRQDWLRQ RI 3LfV 6PLWK DQG )LVFKHU f DQG %LQGHU DQG :DLVV f \LHOGHG IUDFWLRQV ZLWK LQWHUPHGLDWH DQG ZLWK SRODU VROYHQWV WKDW FDXVHG VORZHU GHYHORSPHQWDO UDWHV DQG KLJK PRUWDOLW\ LQ VR\EHDQ ORRSHU >3VHXGRSOXVLD LQFOXGHQV :DONHUf@ DQG +HOLRWKLV ]HD %RGGLH UHVSHFWLYHO\

PAGE 21

6HYHQ SKHQROLF DFLGV ZHUH GHWHFWHG LQ 3, DQG LQ FY f)RUUHVWf JDOOLF SURWRFDWHFKXLF SK\GUR[\EHQ]RLF YDQLOOLF FDIIHLF SFRXPDULF DQG IHUXOLF 7KH SRWHQWLDOO\ DQWLELRWLF FDIIHLF DQG IHUXOLF DFLGV RFFXUUHG DW KLJKHU FRQFHQWUDWLRQV LQ WKH UHVLVWDQW 3, WKDQ LQ WKH VXVFHSWLEOH FXOWLYDU ,Q ERWK W\SHV RI SODQWV FRQFHQWUDWLRQV ZHUH KLJKHU LQ LQMXUHG WKDQ LQ XQLQMXUHG WLVVXH +DUGLQ f ,VRIODYRQRLG SK\WRDOH[LQV LQ VR\EHDQ FRW\OHGRQV DUH SRWHQW IHHGLQJ GHWHUUHQWV IRU WKH 0H[LFDQ EHDQ EHHWOH +DUW HW DO f 7KHVH SK\WRDOH[LQV DUH ELRFKHPLFDOO\ UHODWHG WR WKH SKHQROLFV PHQWLRQHG DERYH WKURXJK WKH VKLNLPLF SDWKZD\ Sf $OWKRXJK D JUHDW GHDO RI NQRZOHGJH KDV EHHQ JDLQHG WKHUH VWLOO UHPDLQ PDQ\ XQDQVZHUHG TXHVWLRQV DV WR ZKDW FRPSRQHQWV FRPSULVH WKH YDULRXV PHFKDQLVPV RI UHVLVWDQFH $ PRGHUDWH DPRXQW RI UHVHDUFK KDV EHHQ FRQGXFWHG LQ WKH DUHD RI UHVLVWDQFH LQ VR\EHDQ WR LQVHFWV DQG PXFK SURJUHVV KDV EHHQ PDGH LQ FXOWLYDU GHYHORSPHQW +RZHYHU WKHUH KDV EHHQ YHU\ OLWWOH LQIRUPDWLRQ SXEOLVKHG RQ WKH LQKHULWDQFH RI UHVLVWDQFH WR LQVHFWV LQ VR\EHDQ ,Q WKHLU ZRUN ZLWK WKH 0H[LFDQ EHDQ EHHWOH 6LVVRQ HW DO f K\SRWKHVL]HG WKDW UHVLVWDQFH ZDV LQKHULWHG TXDQWLWDWLYHO\ DQG WKDW WKH QXPEHU RI PDMRU JHQHV LQYROYHG ZDV VPDOO SHUKDSV WZR RU WKUHH PDMRU JHQHV ,Q IXUWKHU VWXGLHV ZLWK WKH 0H[LFDQ EHDQ EHHWOH 0HEUDKWX HW DO f FRQFOXGHG WKDW UHVLVWDQFH DSSHDUHG WR EH SULPDULO\ DQ H[SUHVVLRQ RI DGGLWLYH JHQHWLF YDULDQFH .LOHQ HW DO f LQ WKHLU VWXGLHV ZLWK WKH VR\EHDQ ORRSHU VXJJHVWHG WKDW VXVFHSWLELOLW\ ZDV SDUWLDOO\ GRPLQDQW ZLWK D IHZ PDMRU JHQHV FRQGLWLRQLQJ UHVLVWDQFH ,Q WKHLU VWXGLHV ZLWK 3, 3, DQG 3, DQG WKH ) SODQWV GHULYHG IURP LQWHUFURVVHV DPRQJ WKH WKUHH 3LfV /DPEHUW DQG .LOHQ f VXJJHVWHG WKDW JHQHWLF UHVLVWDQFH LQ 3, GLIIHUHG IURP WKDW RSHUDWLQJ LQ WKH RWKHU WZR OLQHV 7KLV ZDV IXUWKHU VXSSRUWHG E\ WKHLU ZRUN ZLWK WKH 3LfV DQG WKH )

PAGE 22

OLQHV GHULYHG IURP WKH LQWHUFURVVHV DPRQJ WKH 3LfV 7KH\ K\SRWKHVL]HG WKDW HDFK 3, FRQWDLQHG D UHVLVWDQFH JHQH WKDW GLIIHUHG IURP WKRVH RI WKH RWKHU OLQHV 7R GHVLJQ D EHWWHU EUHHGLQJ SURJUDP WR GHYHORS FXOWLYDUV UHVLVWDQW WR WKH VR\EHDQ ORRSHU XWLOL]LQJ EUHHGLQJ OLQHV WKDW GHULYH WKHLU LQVHFW UHVLVWDQFH IURP 3, LW ZRXOG EH DGYDQWDJHRXV LI WKH PRGH RI LQKHULWDQFH RI WKLV WUDLW ZHUH NQRZQ

PAGE 23

&+$37(5 (9$/8$7,21 2) 5$7,1* 0(7+2'6 ,QWURGXFWLRQ 7KURXJKRXW WKH SDVW \HDUV WKHUH KDYH EHHQ DGYDQFHV LQ WKH LGHQWLILFDWLRQ DQG GHYHORSPHQW RI VR\EHDQ EUHHGLQJ OLQHV UHVLVWDQW WR SK\WRSKDJRXV LQVHFWV HVSHFLDOO\ /HSLGRSWHURXV VSHFLHV DQG WKH 0H[LFDQ EHDQ EHHWOH f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f 6KRNHV HW DO f WHVWHG WKH UHOLDELOLW\ RI GLVHDVH DVVHVVPHQW SURFHGXUHV ZLWK ODWH OHDIVSRW >&HUFRVSRULGLXP SHUVRQDWXP %HUN DQG &XUWf 'HLJKWRQ@ LQ SHDQXW $UDFKLV K\SRJDHD /f DQG FRQFOXGHG

PAGE 24

WKDW SODQW SDWKRORJLVWV ZRXOG EHQHILW LI D VWDQGDUG UDWLQJ V\VWHP IRU IROLDU GLVHDVHV RI SHDQXW FRXOG EH DGRSWHG DV KDV EHHQ GRQH LQ RWKHU FURSV ,Q D VSHFLDO UHSRUW %HUJHU f VWDWHG WKDW D GLVHDVH DVVHVVPHQW SURFHGXUH VKRXOG KDYH WKH IROORZLQJ EDVLF UHTXLUHPHQWV HDV\ WR OHDUQ TXLFN DQG HIILFLHQW WR XVH DSSOLFDEOH DFURVV D EURDG VSHFWUXP RI FRQGLWLRQV DQG EH DFFXUDWH SUHFLVH DQG UHSURGXFLEOH +H FRQFOXGHG WKDW WKH JUHDWHU WKH HPSKDVLV SODFHG RQ WKH DVVHVVPHQW PHWKRG WKH KLJKHU WKH TXDOLW\ RI LQIRUPDWLRQ REWDLQHG E\ WKH PHWKRG ,Q .RJDQ DQG 7XPLSVHHG f GHVFULEHG WKH YDULRXV PHWKRGV RI TXDQWLWDWLYH DQG TXDOLWDWLYH DVVHVVPHQWV RI GHIROLDWLRQ WKDW DUH FXUUHQWO\ XVHG E\ UHVHDUFKHUV ZRUNLQJ ZLWK SK\WRSKDJRXV LQVHFWV 7KHVH PHWKRGV DUH Lf SKRWRHOHFWULF RU HOHFWURQLF PHDVXUH DFWXDO OHDI DUHDf LLf JUDYLPHWULF UHJUHVVLQJ NQRZQ OHDI DUHDV RQ WKH FRUUHVSRQGLQJ IUHVK RU GU\ ZHLJKW RI WKHVH DUHDV LLLf YROXPHWULF YROXPH GLVSODFHPHQW F\OLQGHU LV XVHG WR PHDVXUH ELRPDVVf LYf SODQLPHWULF PHFKDQLFDO PHDVXUHPHQW RI DUHDf Yf JHRPHWULF HVWLPDWHV WRWDO QRQGHIROLDWHG OHDI DUHD VR WKDW SHUFHQW GHIROLDWLRQ FDQ EH FDOFXODWHG ZKHQ XVHG LQ FRQMXQFWLRQ ZLWK D OHDI DUHD PHWHUf DQG YLf YLVXDO HVWLPDWHV SURYLGH UDSLG HVWLPDWH RI GHIROLDWLRQf $OWKRXJK GHVFULSWLRQV RI WKH PHWKRGV ZHUH GHWDLOHG E\ .RJDQ DQG 7XPLSVHHG QR VWDWLVWLFDO FRPSDULVRQV ZHUH PDGH DPRQJ PHWKRGV 6RPH DXWKRUV f UHIHU WR TXDOLWDWLYH DVVHVVPHQWV DV UHODWLYH HVWLPDWHV DQG TXDQWLWDWLYH DVVHVVPHQWV DV DEVROXWH HVWLPDWHV 5HODWLYH HVWLPDWHV GR QRW WUDQVODWH GLUHFWO\ LQWR WKH DPRXQW RI GHIROLDWLRQ ZKHUHDV DEVROXWH HVWLPDWHV DOORZ IRU GLUHFW WUDQVODWLRQ LQWR WKH DPRXQW RI GHIROLDWLRQ FDXVHG E\ SK\WRSKDJRXV LQVHFWV 7KH PDMRU GUDZEDFN RI DEVROXWH HVWLPDWHV LV WKDW WKH\ DUH XVXDOO\ WLPH FRQVXPLQJ DQG FRVWO\ WR PDNH

PAGE 25

&RQYHUVHO\ UHODWLYH HVWLPDWHV FDQ EH WLPHO\ DQG DUH LQH[SHQVLYH ZKLFK LV ZK\ WKHVH W\SHV RI HVWLPDWHV DUH PRUH ZLGHO\ XVHG E\ UHVHDUFKHUV f )XQGHUEXUN HW DO f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f WKH SDUWLWLRQHG SODQW DYHUDJH YLVXDO UDWLQJ $9f DQG PHDVXUHG DYHUDJH OHDI DUHD RI WKH SDUWLWLRQHG SODQW $/f :3 YDOXHV ZHUH EDVHG RQ D OWR VFDOH GLYLGHG LQWR b LQFUHPHQWV ZLWK D VFRUH RI VLJQLI\LQJ GHIROLDWLRQ RI WR b DQG D VFRUH RI VLJQLI\LQJ GHIROLDWLRQ RI WR b $OWKRXJK RWKHU VFDOHV VXFK DV D ORJDULWKPLF VFDOH f PD\ EH EHWWHU DW HVWLPDWLQJ GHIROLDWLRQ WKH OWR VFDOH ZDV FKRVHQ VLQFH LW LV PRVW FRPPRQO\ XVHG E\ SDUWLFLSDQWV LQ WKH 5HJLRQDO +RVW 3ODQW 5HVLVWDQFH 7HVW ZKLFK LV VSRQVRUHG DQQXDOO\ E\ WKH 6RXWKHUQ 5HJLRQDO ,QIRUPDWLRQ ([FKDQJH *URXS 65,(* f :LWK WKH $9 DQG $/ PHWKRGV WKH FDQRSLHV ZHUH YLVXDOO\ SDUWLWLRQHG LQWR WKUHH VHFWLRQV WRS PLGGOH ERWWRPf HQDEOLQJ D PRUH GHWDLOHG DVVHVVPHQW RI VR\EHDQ ORRSHU GHIROLDWLRQ RQ HDFK SODQW 7KH SDUWLWLRQLQJ RI WKH SODQWV ZDV EDVHG RQ D SURFHGXUH WKDW ZDV GHYHORSHG E\ 3ODXW DQG %HUJHU f WKDW ZDV FLWHG E\ 6KRNHV HW DO f )RU $9

PAGE 26

HVWLPDWHV HDFK VHFWLRQ RI WKH SODQW ZDV VFRUHG XVLQJ WKH VDPH VFDOH DV WKH :3 ZLWK WKH WKUHH VFRUHV DYHUDJHG WR REWDLQ WKH $9 ,Q D /,&25 /, /L&RU ,QF /LQFROQ 1( f SRUWDEOH OHDI DUHD PHWHU ZDV XVHG WR PHDVXUH WKH PLGGOH OHDIOHW RI D UHSUHVHQWDWLYH WULIROLDWH IURP HDFK RI WKH VHFWLRQV DQG WKH WKUHH PHDVXUHPHQWV ZHUH DYHUDJHG WR REWDLQ WKH $/ ,Q WKH $/ PHWKRG RQO\ DOORZHG DQ LQGLUHFW FRPSDULVRQ WR WKH RWKHU WZR PHWKRGV WKHUHIRUH WKH $/ ZDV PRGLILHG LQ WR SHUPLW D GLUHFW FRPSDULVRQ ZLWK $9 DQG :3 ,Q WKH $/ PHWKRG D UHSUHVHQWDWLYH WULIROLRODWH OHDI IURP HDFK VHFWLRQ RI WKH SDUWLWLRQHG SODQW ZDV UHPRYHG DQG WDNHQ LQWR D ODERUDWRU\ WR GHWHUPLQH WKH DUHD RI HDFK VDPSOH 7KH SRUWDEOH OHDI DUHD PHWHU ZDV XVHG WR GHWHUPLQH WKH UHPDLQLQJ SRUWLRQ RI HDFK GHIROLDWHG OHDIOHW '()f RI HDFK WULIROLRODWH )ROORZLQJ WKH SURFHGXUH GHVFULEHG E\ .RJDQ DQG 7XPLSVHHG f HDFK OHDIOHW RI WKH WULIROLRODWHV ZDV WKHQ SODFHG RQ D SKRWRFRS\ PDFKLQH DQG D FRS\ ZDV PDGH ([DPSOHV RI OHDIOHW SKRWRFRSLHV DUH VKRZQ LQ )LJXUH 7KHVH VLOKRXHWWHV ZHUH WKHQ FXW RXW WR UHSUHVHQW WKH WULIROLRODWHV LQ D QRQGHIROLDWHG VWDWH DQG WKH DUHD 121'()f IRU HDFK ZDV GHWHUPLQHG DV PHQWLRQHG DERYH 7KH WZR YDOXHV REWDLQHG IRU HDFK WULIROLRODWH ZHUH XVHG LQ WKH IROORZLQJ IRUPXOD WR GHWHUPLQH WKH SHUFHQW GHIROLDWLRQ b GHIf 121'() '()f121'() r b GHI 7KH SHUFHQW GHIROLDWLRQ IRU HDFK RI WKH WKUHH WULIROLRODWHV IURP HDFK SODQW ZDV WKHQ FRQYHUWHG WR WKH VDPH OWR VFDOH XVHG E\ WKH RWKHU WZR PHWKRGV DQG WKH YDOXHV DYHUDJHG DOORZLQJ GLUHFW FRPSDULVRQV WR EH PDGH DPRQJ WKH WKUHH PHWKRGV 7KH SODQWV XVHG LQ WKH HYDOXDWLRQ RI WKH WKUHH PHWKRGV ZHUH SDUW RI DQ LQKHULWDQFH VWXG\ WKDW ZDV FRQGXFWHG LQ ILHOG FDJHV f LQ DQG DW WKH 0LVVLVVLSSL 6WDWH

PAGE 27

)LJXUH ([DPSOHV RI SKRWRFRSLHG OHDIOHWV XVHG LQ GHWHUPLQLQJ WKH b GHIROLDWLRQ LQ WKH $9 OHDI DUHD RI SDUWLWLRQHG SODQWf PHWKRG

PAGE 28

8QLYHUVLW\ 'HOWD 5HVHDUFK DQG ([WHQVLRQ &HQWHU 6WRQHYLOOH 0LVVLVVLSSL ,Q WKH ILUVW \HDU RI WKH VWXG\ VHHG ZHUH SODQWHG RQ 0D\ ,Q VHHG ZHUH SODQWHG RQ 0D\ DQG 0D\ LQ FDJH DQG RQ 0D\ LQ FDJH 5DLQIDOO RQ 0D\ SUHYHQWHG FRPSOHWLRQ RI SODQWLQJ RQ WKH VDPH GD\ LQ FDJH 7R FRPSDUH PHWKRGV LQ WHQ VR\EHDQ SODQWV ZHUH FKRVHQ UDQGRPO\ IURP HDFK RI EDFNFURVV IDPLOLHV IRU D WRWDO RI SODQWV )LIW\ SODQWV LQ HDFK RI WZR FDJHV ZHUH UDQGRPO\ FKRVHQ XVLQJ 352& 3/$1 RI 3&6$6 f IRU XVH LQ WKH HYDOXDWLRQ RI WKH WKUHH PHWKRGV LQ 7KHVH SODQWV ZHUH FKRVHQ SULRU WR UHOHDVH RI VR\EHDQ ORRSHU PRWKV LQ ERWK \HDUV 7KH WKUHHGD\ROG VR\EHDQ ORRSHU PRWKV ZHUH UHOHDVHG RQ -XQH ZKHQ SODQWV ZHUH DW WKH 9 WR 9 VWDJH RI SODQW JURZWK f /DUYDH IURP KDWFKHG HJJV ZHUH DOORZHG WR IHHG IUHHO\ IRU DSSUR[LPDWHO\ GD\V DW ZKLFK WLPH f&HQWHQQLDOf f WKH VXVFHSWLEOH FKHFN ZDV HVWLPDWHG WR KDYH D UDWLQJ RI QLQH DFFRUGLQJ WR WKH :3 PHWKRG 7KH DUHD ZLWKLQ WKH FDJH ZDV WKHQ VSUD\HG ZLWK PHWKRP\O >6PHWK\O1PHWK\O FDUEDPR\Of R[\fWKLRDFHWLPLGDWH@ DW NJ KDn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

PAGE 29

LQ FDJH DQG EHJLQQLQJ 5 LQ FDJH WKH SODQWV LQ ERWK FDJHV ZHUH VSUD\HG ZLWK PHWKRP\O DW NJ KD WR NLOO WKH ODUYDH 7KH IROORZLQJ GD\ HDFK RI WKH SODQWV ZDV UDWHG XVLQJ WKH :3 DQG $9 PHWKRGV DQG VDPSOHV ZHUH FROOHFWHG IRU WKH $/ 'DWD GHULYHG IURP HDFK UDWLQJ PHWKRG IRU WKH SODQWV LQ DQG WKH SODQWV LQ ZHUH VXEMHFWHG WR DQDO\WLFDO SURFHGXUHV LQ 3&6$6 f $ FRUUHODWLRQ DQDO\VLV ZDV UXQ WR GHWHUPLQH WKH FRUUHODWLRQ FRHIILFLHQW Uf EHWZHHQ :3 DQG $/ $9 DQG :3 DQG $/ DQG :3 UDWLQJ PHWKRGV HDFK \HDU $ SDLUHG GLIIHUHQFH FRPSDULVRQ f ZDV PDGH EHWZHHQ :3 DQG $9 PHWKRGV LQ DQG EHWZHHQ :3 DQG $/ :3 DQG $9 DQG $/ DQG $9 PHWKRGV LQ 0F1HPDUfV WHVW IRU FRQFRUGDQFH f ZDV XVHG WR WHVW IRU VLJQLILFDQW FKDQJHV EHWZHHQ :3 DQG $/ LQ DQG EHWZHHQ :3 DQG $/ :3 DQG $9 DQG $/ DQG $9 PHWKRGV LQ 7KLV DQDO\VLV ZDV FRQGXFWHG WR GHWHUPLQH LI WKH PHWKRGV GLIIHUHG LQ WKHLU HVWLPDWHV RI GHIROLDWLRQ 3UHFLVLRQ NQRZQ DV WKH UHODWLYH YDULDWLRQ 59f RI WKH WZR UHODWLYH PHWKRGV IRU ERWK \HDUV DQG WKH DEVROXWH PHWKRG LQ ZHUH FDOFXODWHG IRU HDFK DV IROORZV 59 6(-(f ZKHUH 6( LV WKH VWDQGDUG HUURU RI WKH PHDQ DQG ; LV WKH PHDQ 7KLV UHODWLRQVKLS GHVFULEHG E\ 5XHVLQN f DQG 3HGLJR f LV FRQVLGHUHG XVHIXO LQ HVWLPDWLQJ SUHFLVLRQ RI D VDPSOLQJ PHWKRG E\ HQWRPRORJLVWV 5HVXOWV DQG 'LVFXVVLRQ 7KH VFRUHV IRU HDFK RI WKH WKUHH UDWLQJ PHWKRGV E\ SODQW DUH SUHVHQWHG LQ $SSHQGL[ $ 6FRUHV IRU $9 DQG :3 ZHUH EDVHG RQ D OWR VFDOH ZLWK ORZHU QXPEHUV

PAGE 30

LQGLFDWLQJ OHVV GHIROLDWLRQ 7KH PHDVXUHPHQWV IRU $/ LQ DUH SUHVHQWHG DV FP RI UHPDLQLQJ WLVVXH ,QGLYLGXDO SODQW GDWD IRU HDFK RI WKH WKUHH PHWKRGV LQ DUH SUHVHQWHG LQ $SSHQGL[ % 8QOLNH GDWD IRU DOO WKUHH PHWKRGV ZHUH EDVHG RQ D OWR VFDOH 0HDQV DQG VWDQGDUG GHYLDWLRQV IRU ERWK \HDUV DUH DOVR UHSRUWHG LQ $SSHQGLFHV $ DQG % )UHTXHQF\ GLVWULEXWLRQV RI WKH :3 DQG $9 UDWLQJ PHWKRGV DUH VKRZQ LQ )LJXUH 7KH $/ PHWKRG ZDV RPLWWHG IURP WKLV JUDSK EHFDXVH LW PHDVXUHG WKH UHPDLQLQJ XQGHIROLDWHG WLVVXH LQ FP ZKHUHDV WKH $9 DQG :3 PHWKRGV ZHUH XVHG WR HVWLPDWH SHUFHQW GHIROLDWLRQ RQ D OWR VFDOH )UHTXHQF\ GLVWULEXWLRQV RI WKH $9 DQG :3 PHWKRGV VXJJHVW WKDW HLWKHU WKH HYDOXDWRU XQGHUHVWLPDWHG GHIROLDWLRQ E\ WKH :3 PHWKRG RU RYHUHVWLPDWHG GHIROLDWLRQ E\ WKH $9 PHWKRG :LWKRXW DQ DEVROXWH HVWLPDWH RI GHIROLDWLRQ QR VXEVWDQWLDO FRQFOXVLRQ FDQ EH PDGH DV WR WKH UHOLDELOLW\ RI $9 RU :3 WR DFFXUDWHO\ DVVHVV GHIROLDWLRQ 7KH IUHTXHQF\ GLVWULEXWLRQV RI DOO WKUHH PHWKRGV IRU ERWK FDJHV FRPELQHG DUH SUHVHQWHG LQ )LJXUH 'XH WR WKH PRGLILFDWLRQV LQ WKH $/ PHWKRG LW ZDV SRVVLEOH WR PDNH GLUHFW FRPSDULVRQV EHWZHHQ $/ DQG WKH YLVXDO UDWLQJ PHWKRGV 7KH HYDOXDWRU RYHUHVWLPDWHG GHIROLDWLRQ WR REWDLQ WRR IHZ SODQWV EHORZ b RU LQ WKH OWR VFDOH ZKHQ XVLQJ HLWKHU UHODWLYH PHWKRG :3 RU $9 DV FRPSDUHG WR $/ WKH DEVROXWH PHWKRG $OVR WKH HYDOXDWRU KDG D WHQGHQF\ WR RYHUHVWLPDWH WKH DPRXQW RI GHIROLDWLRQ RYHU b RU LQ WKH OWR VFDOH ZKHQ XVLQJ HLWKHU UHODWLYH PHWKRG DV FRPSDUHG WR $/ 2QO\ LQ FODVVHV DQG ZDV WKH HYDOXDWRUfV HVWLPDWH RI GHIROLDWLRQ ZLWK WKH :3 DQG $9 PHWKRGV VLPLODU WR WKDW RI WKH $/ 7KH SUREOHP RI RYHUHVWLPDWLQJ PLJKW EH DYRLGHG LQ

PAGE 31

1XPEHU RI 3, DQWV 'HIROLDWLRQ 6FRUHV )LJXUH IUHTXHQF\ GLVWULEXWLRQ RI GHIROLDWLRQ HVWLPDWHV E\ WKH ZKROH SODQW DQG SDUWLWLRQHG SODQW DYHUDJH YLVXDO UDWLQJ PHWKRGV 'HIROLDWLRQ VFRUHV EDVHG RQ OWR VFDOH ZKHUH WR b GHIROLDWLRQ WR b WR b WR b WR b WR b WR b WR b WR b WR b

PAGE 32

1XPEHU RI 3ODQWV 'HIROLDWLRQ 6FRUHV )LJXUH IUHTXHQF\ GLVWULEXWLRQ RI GHIROLDWLRQ HVWLPDWHV E\ WKH ZKROH SODQW SDUWLWLRQHG SODQW DYHUDJH YLVXDO DQG PHDVXUHG DYHUDJH OHDI DUHD RI WKH SDUWLWLRQHG SODQW UDWLQJ PHWKRGV 'HIROLDWLRQ VFRUHV EDVHG RQ OWR VFDOH ZKHUH WR b GHIROLDWLRQ WR b WR b WR b WR b WR b WR b WR b WR b WR b

PAGE 33

WKH IXWXUH LI D ORJDULWKPLF UDWLQJ VFDOH VXFK DV WKH +RUVHIDOO%DUUDWW VFDOH WR DVVHVV GLVHDVH f ZHUH XVHG WR HVWLPDWH GHIROLDWLRQ $ ORJDULWKPLF VFDOH PD\ KDYH DGYDQWDJHV IRU DVVHVVPHQW RI GHIROLDWLRQ EHFDXVH WKH YLVXDO DFXLW\ RI WKH KXPDQ H\H LV PRUH DGDSWHG WR WKH LGHQWLILFDWLRQ RI VPDOO GLIIHUHQFHV DW HDFK HQG RI WKH VFDOH 7KH SODQW VFRUHV RI WKH $9 DQG WKH :3 UDWLQJ PHWKRGV LQ ZHUH KLJKO\ FRUUHODWHG U S f $/ PHDVXUHPHQWV ZHUH LQ FP RI QRQGHIROLDWHG OHDI DUHD UHPDLQLQJ WKHUHIRUH QRW D PHDVXUH RI GHIROLDWLRQ $V WKHUH ZDV QR DEVROXWH PHDVXUH RI GHIROLDWLRQ LW ZDV QRW SRVVLEOH WR PDNH IXUWKHU FRPSDULVRQV ZLWK $9 DQG :3 ,Q WKH VFRUHV IRU FDJH IURP WKH $9 DQG :3 PHWKRGV ZHUH VLJQLILFDQWO\ FRUUHODWHG ZLWK WKH $/ PHWKRG U S DQG U S UHVSHFWLYHO\f 7KH $9 DQG :3 PHWKRGV FRUUHODWHG ZHOO ZLWK HDFK RWKHU U S f %\ VTXDULQJ HDFK RI WKH FRUUHODWLRQ FRHIILFLHQWV Uf ERWK WKH :3 YV $9 DQG :3 YV $/ UHODWLRQVKLSV KDG U YDOXHV RI PHDQLQJ b RI WKH YDULDQFH LQ RQH UDWLQJ PHWKRG FDQ EH H[SODLQHG E\ WKH YDULDWLRQ LQ WKH RWKHU 7KH $9 YV $/ FRUUHODWLRQ KDG D U YDOXH RI LQGLFDWLQJ WKDW b RI WKH YDULDQFH RI WKH $9 PHWKRG FDQ EH H[SODLQHG E\ WKH YDULDWLRQ LQ WKH $/ PHWKRG RU YLFH YHUVDf $OWKRXJK WKH UHVXOWV IURP UDWLQJV RQ SODQWV LQ FDJH ZHUH VLPLODU FRUUHODWLRQV ZHUH QRW DV KLJK DV WKRVH LQ FDJH 6FRUHV IURP WKH $9 DQG :3 PHWKRGV ZHUH VLJQLILFDQWO\ FRUUHODWHG ZLWK WKH $/ PHWKRG U S DQG U S UHVSHFWLYHO\f 7KH $9 DQG :3 PHWKRGV ZHUH VLJQLILFDQWO\ FRUUHODWHG U S f ZLWK HDFK RWKHU &DOFXODWHG U YDOXHV ZHUH :3 YV $9 U :3 YV

PAGE 34

$/ U DQG $9 YV $/ U 7KHVH YDOXHV DUH ORZHU WKDQ WKRVH IRXQG ZLWK WKH FDJH GDWD 'LIIHUHQFHV PD\ EH DWWULEXWHG WR GLIIHUHQFHV LQ WKH SODQW JURZWK DW WKH WLPH RI UDWLQJ RU WR HYDOXDWRU HUURU 7KH FDJH UDWLQJV ZHUH PDGH LQ WKH UDLQ ZKLFK KLQGHUHG WKH HYDOXDWRUfV DELOLW\ WR DFFXUDWHO\ HVWLPDWH GHIROLDWLRQ ,W ZDV FRQVLGHUHG PRUH SUXGHQW WR PDNH UDWLQJV LQ WKH UDLQ WKDQ WR DWWHPSW WR GLVUHJDUG WKH HQVXLQJ UHJURZWK ZKLFK ZDV LQFUHDVLQJ UDSLGO\ GXULQJ D SURORQJHG UDLQ\ SHULRG :KHQ GDWD IRU ERWK FDJHV ZHUH SRROHG IRU DQDO\VLV VFRUHV IRU WKH $9 DQG :3 PHWKRGV ZHUH VLJQLILFDQWO\ FRUUHODWHG ZLWK WKH $/ PHWKRG U S DQG U S UHVSHFWLYHO\f 7KH $9 DQG :3 PHWKRGV FRUUHODWHG ZHOO ZLWK HDFK RWKHU U S f DV LQ WKH FDJHV WUHDWHG VHSDUDWHO\ 7KH IROORZLQJ U YDOXHV ZHUH REWDLQHG :3 YV $9 U :3 YV $/ U DQG $9 YV $/ U WKXV DSSUR[LPDWHO\ b s bf RI WKH YDULDWLRQ LQ RQH UDWLQJ PHWKRG FRXOG EH DWWULEXWHG WR WKH YDULDWLRQ LQ WKH RWKHU UDWLQJ PHWKRG ,Q D VLPLODU VWXG\ D UHODWLYH PHWKRG DQG DQ DEVROXWH PHWKRG ZHUH FRPSDUHG XVLQJ WKHLU HVWLPDWHV RI GHIROLDWLRQ RI ILHOG SORWV LQIHVWHG SULPDULO\ b RI WRWDO LQVHFW SRSXODWLRQf ZLWK WKH YHOYHWEHDQ FDWHUSLOODU *5 %RZHUV SHUVRQDO FRPPXQLFDWLRQf $ OWR VFDOH ZLWK VLJQLI\LQJ WKH OHDVW GDPDJH ZDV XVHG DV WKH UHODWLYH PHWKRG 7KH DEVROXWH PHWKRG IROORZHG D SURFHGXUH GHVFULEHG E\ .RJDQ DQG .XKOPDQ f LQ ZKLFK OHDIOHWV HDFK IURP WKH WRS PLGGOH DQG ERWWRP WKLUGV RI SODQWV IRU D WRWDO RI f ZHUH VDPSOHG IURP SODQWV ORFDWHG LQ WKH FHQWHU WZR URZV RI D IRXUURZ SORW 7KHVH ZHUH WKHQ FRPSDUHG WR GLDJUDPV RI GHIROLDWHG OHDYHV WR GHWHUPLQH WKH GHIROLDWLRQ HVWLPDWH RI HDFK VDPSOH 7KH WZR PHWKRGV ZHUH VLJQLILFDQWO\ FRUUHODWHG U

PAGE 35

S %RZHUf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f DV ZLWK WKH $/ PHWKRG WKH DEVROXWH PHWKRGf 7R FRPSDUH WKH DFFXUDF\ RI WKH HYDOXDWRU XVLQJ WKH WZR UHODWLYH PHWKRGV DQG WKH DEVROXWH PHWKRG DGGLWLRQDO DQDO\VHV ZHUH XQGHUWDNHQ :LWK WKH 352& 0($16 SURFHGXUH RI 3&6$6 D SDLUHG GLIIHUHQFH FRPSDULVRQ ZDV PDGH EHWZHHQ WKH :3 DQG $9 PHWKRGV LQ $ WWHVW ZDV SHUIRUPHG WR WHVW WKH HTXDOLW\ RI WKH WZR PHWKRGV EDVHG RQ WKH GLIIHUHQFHV EHWZHHQ WKH WZR PHWKRGV IRU HDFK SODQW LQ WKH VWXG\ 7KH PHDQ GLIIHUHQFH RI EHWZHHQ WKH :3 DQG $9 PHWKRGV ZDV KLJKO\ VLJQLILFDQW S PHDQLQJ WKH :3 PHWKRG \LHOGHG ORZHU DYHUDJH SHUFHQW GHIROLDWLRQ WKDQ WKH $9 PHWKRG +RZHYHU ZLWKRXW VRPH PHDVXUH RI WKH DFWXDO RU DEVROXWH GHIROLDWLRQ DFURVV WKH SODQWV WKHUH ZDV QR ZD\ WR GHWHUPLQH LI WKH HYDOXDWRU FRXOG REWDLQ DQ DFFXUDWH DVVHVVPHQW RI WKH WUXH GHIROLDWLRQ RI D SODQW ZLWK HLWKHU PHWKRG $ SDLUHG GLIIHUHQFH FRPSDULVRQ ZDV DOVR PDGH DPRQJ WKH WKUHH UDWLQJ PHWKRGV LQ 7KH :3 DQG $9 PHWKRGV KDG D PHDQ GLIIHUHQFH RI S ZKLFK ZDV VRPHZKDW VLPLODU LQ PDJQLWXGH WR WKH PHDQ GLIIHUHQFH FDOFXODWHG LQ WKH VWXG\

PAGE 36

7KH PHDQ GLIIHUHQFHV s VWDQGDUG GHYLDWLRQ LQYROYLQJ WKH $/ PHWKRG ZHUH s IRU :3 $/ DQG s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f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f DQG IROORZ WKH SURWRFRO RI WKH $/ PHWKRG DIWHU FXWWLQJ RXW QHFURWLF WLVVXH ,Q WKH H[SHULPHQW SURYLVLRQV ZHUH QRW PDGH IRU GHVWUXFWLYH VDPSOLQJ WKHUHIRUH LW ZDV QRW SUDFWLFHG 0F1HPDUfV WHVW IRU FRQFRUGDQFH ZDV XVHG WR GHWHUPLQH LI WKH FDWHJRUL]DWLRQ RI SODQWV KDYLQJ OHVV WKDQ RU HTXDO WR b GHIROLDWLRQ DQG RI SODQWV KDYLQJ JUHDWHU WKDQ b GHIROLDWLRQ GLIIHUHG ZLWK WKH :3 DQG $9 PHWKRGV ,Q RI WKH RI WKH SODQWV

PAGE 37

ZHUH FODVVLILHG ZLWK WKH :3 PHWKRG DV KDYLQJ DW PRVW b GHIROLDWLRQ ZKLOH ZLWK WKH $9 PHWKRG WKHVH VDPH SODQWV ZHUH UDWHG DV KDYLQJ JUHDWHU WKDQ b GHIROLDWLRQ 7DEOH f 7KLV SHUFHQWDJH bf RI SODQWV ZDV VLJQLILFDQWO\ GLIIHUHQW IURP ]HUR ; S VXSSRUWLQJ WKH HDUOLHU FRQWHQWLRQ )LJXUH f WKDW HLWKHU WKH :3 PHWKRG XQGHUHVWLPDWHG GHIROLDWLRQ RU WKH $9 PHWKRG RYHUHVWLPDWHG GHIROLDWLRQ :LWKRXW WKH DEVROXWH PHWKRG IRU UHIHUHQFH LW ZDV QRW SRVVLEOH WR GHWHUPLQH ZKLFK PHWKRG ZDV PRUH DFFXUDWH $OO WKUHH PHWKRGV ZHUH FRPSDUHG LQ WR GHWHUPLQH ZKHWKHU VLJQLILFDQW GLIIHUHQFHV DPRQJ WKH PHWKRGV H[LVWHG 7KH UHVXOWV IURP 0F1HPDUfV WHVW IRU FRQFRUGDQFH RI WKH GDWD DUH VKRZQ LQ 7DEOH 7KH WZR UHODWLYH PHWKRGV ZHUH VLJQLILFDQWO\ GLIIHUHQW ; S f ZLWK RI WKH SODQWV EHLQJ FODVVLILHG GLIIHUHQWO\ E\ :3 WKDQ $9 7KH UHODWLYH PHWKRGV ZHUH ERWK VLJQLILFDQWO\ GLIIHUHQW IURP WKH DEVROXWH PHWKRG :3 YV $/ ; S DQG $9 YV $/ ; S f LQ WKH HVWLPDWH RI GHIROLDWLRQ RQ D SODQWE\SODQW EDVLV :LWK ERWK UHODWLYH PHWKRGV WKH HYDOXDWRU KDG D WHQGHQF\ WR RYHUHVWLPDWH GHIROLDWLRQ DV FRPSDUHG WR WKH DEVROXWH PHWKRG :LWK WKH $9 PHWKRG RI SODQWV ZHUH FODVVLILHG DV b GHIROLDWHG ZKLOH $/ FODVVLILHG WKHVH VDPH SODQWV DV b ZKHUHDV ZLWK WKH :3 PHWKRG RQO\ RI SODQWV ZHUH FODVVLILHG b DV FRPSDUHG WR WKH $/ PHWKRG FODVVLI\LQJ WKHP DV b 7KLV LQGLFDWHV WKDW WKH HYDOXDWRU ZDV DEOH WR HVWLPDWH GHIROLDWLRQ FORVHU WR WKH DEVROXWH PHWKRG ZLWK WKH :3 PHWKRG DV ZDV VHHQ LQ WKH SDLUHG GLIIHUHQFH FRPSDULVRQ

PAGE 38

7DEOH 0F1HPDU WHVW IRU VLJQLILFDQW GLIIHUHQFHV EHWZHHQ WKH :3 ZKROH SODQW YLVXDOf DQG $9 SDUWLWLRQHG SODQW YLVXDOf PHWKRGV LQ DQG EHWZHHQ WKH :3 DQG $9 :3 DQG $/ OHDI DUHD RI WKH SDUWLWLRQHG SODQWf DQG $9 DQG $/ PHWKRGV LQ $9 :3 W W 7RWDOV ; 7RWDOV rrr :3 $9 ! 7RWDOV rrr :3 $/ ! 7RWDOV fr $9 $/ ! 7RWDOV rrr rrr 6LJQLILFDQW DW WKH OHYHO RI SUREDELOLW\ W 'HIROLDWLRQ UDWLQJV EDVHG RQ D WR VFDOH ZKHUH WR b GHIROLDWLRQ WR b WR b WR b WR b WR b WR b WR b WR b WR b

PAGE 39

59 YDOXHV IRU ZHUH DQG IRU :3 DQG $9 UHVSHFWLYHO\ $FFRUGLQJ WR 3HGLJR f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f 7KH HYDOXDWRU HVWLPDWHG D VOLJKWO\ ORZHU DPRXQW RI GHIROLDWLRQ RQ WKH DYHUDJH ZLWK WKH :3 PHWKRG WKDQ ZLWK WKH $9 7KH UHODWLYH YDULDWLRQ RI WKHVH WZR UHODWLYH PHWKRGV ZHUH ERWK OHVV WKDQ ZKLFK DFFRUGLQJ WR 3HGLJR f LV YHU\ GHVLUDEOH :LWKRXW DQ DGHTXDWH DEVROXWH PHWKRG QR GHWHUPLQDWLRQ FRXOG EH PDGH DV WR ZKLFK UHODWLYH PHWKRG EHVW UHIOHFWHG WKH DFWXDO GHIROLDWLRQ

PAGE 40

$GMXVWPHQWV LQ WKH PHWKRG RI FDOFXODWLQJ SHUFHQW GHIROLDWLRQ E\ WKH $/ PHWKRG IRU DOORZHG IRU GLUHFW HYDOXDWLRQV RI WKH WZR UHODWLYH PHWKRGV $OO FRUUHODWLRQV :3 YV $9 :3 YV $/ DQG $9 YV $/f ZHUH VLJQLILFDQW DW WKH S OHYHO ZKHWKHU WKH FDJH GDWD ZHUH WUHDWHG VHSDUDWHO\ RU SRROHG 7KH WZR UHODWLYH HVWLPDWHV RI GHIROLDWLRQ ZHUH VLJQLILFDQWO\ FRUUHODWHG ZLWK WKH DEVROXWH HVWLPDWH RI GHIROLDWLRQ 7KHVH UHVXOWV DUH VLPLODU WR WKH ILQGLQJV RI D DQRWKHU UHODWLYH YHUVXV DQ DEVROXWH HVWLPDWLRQ RI GHIROLDWLRQ FRQGXFWHG LQ E\ *5 %RZHUV *5 %RZHUV SHUVRQDO FRPPXQLFDWLRQf 3DLUHG FRPSDULVRQV ZHUH PDGH DPRQJ WKH PHWKRGV WR WHVW WKH HTXDOLW\ RI WKH PHWKRGV WR UDWH GHIROLDWLRQ ,Q WKH HVWLPDWHV RI GHIROLDWLRQ E\ DOO WKUHH PHWKRGV ZHUH VLJQLILFDQWO\ GLIIHUHQW 7KHVH ILQGLQJV ZHUH YDOLGDWHG E\ 0F1HPDUfV WHVW IRU FRQFRUGDQFH 7KH :3 GHIROLDWLRQ HVWLPDWHV ZHUH PRUH VLPLODU WR WKH $/ UHVXOWV WKDQ WKH $9 UDWLQJV ZHUH WR $/ %DVHG RQ WKH 59fV RI HDFK PHWKRG WKH SUHFLVLRQ RI WKH $9 DQG :3 PHWKRGV ZDV HVVHQWLDOO\ HTXDO LQ 6LPLODU UHVXOWV ZHUH REWDLQHG LQ ZLWK WKH $9 :3 DQG $/ PHWKRGV ,Q ERWK \HDUV WKH SUHFLVLRQ RI WKH PHWKRGV ZDV HVVHQWLDOO\ HTXDO DQG ZHOO EHORZ WKH YDOXH RI WKDW 3HGLJR f VWDWHV DV EHLQJ GHVLUDEOH 7KHVH UHVXOWV LQGLFDWH WKDW HLWKHU UHODWLYH PHWKRG LV VXLWDEOH IRU HVWLPDWLQJ GHIROLDWLRQ 7KH UHVXOWV RI WKHVH VWXGLHV GHPRQVWUDWH WKDW UHODWLYH HVWLPDWHV RI GHIROLDWLRQ FDQ EH XVHG HIIHFWLYHO\ LQ JHQHWLF VWXGLHV RU URXWLQH VFUHHQLQJV RI DGYDQFHG EUHHGLQJ OLQHV 6LQFH WKH WZR UHODWLYH PHWKRGV \LHOGHG VLPLODU UHVXOWV LW ZRXOG EH DGYDQWDJHRXV WR XVH WKH PRVW HFRQRPLFDO PHWKRG

PAGE 41

&+$37(5 ,1+(5,7$1&( 2) 5(6,67$1&( 72 62<%($1 /223(5 ,1 62<%($1 ,QWURGXFWLRQ 'XULQJ WKH SDVW WKUHH GHFDGHV JUHDW VWULGHV KDYH EHHQ PDGH LQ WKH LGHQWLILFDWLRQ DQG GHYHORSPHQW RI LQVHFWUHVLVWDQW EUHHGLQJ OLQHV RI VR\EHDQ f 7KLV UHVHDUFK KDV OHG WR WKH UHOHDVH RI VHYHUDO JHUPSODVP OLQHV DQG FXOWLYDUV f WKDW DUH UHVLVWDQW WR SK\WRSKDJRXV LQVHFWV $OWKRXJK DGYDQFHV KDYH EHHQ PDGH YHU\ OLWWOH LQIRUPDWLRQ KDV EHHQ SXEOLVKHG RQ WKH LQKHULWDQFH RI UHVLVWDQFH WR IROLDUIHHGLQJ LQVHFWV 6LVVRQ HW DO f UHSRUWHG WKDW UHVLVWDQFH WR WKH 0H[LFDQ EHDQ EHHWOH ZDV TXDQWLWDWLYHO\ LQKHULWHG ZLWK WZR RU WKUHH PDMRU JHQHV LQYROYHG ,Q RWKHU LQKHULWDQFH VWXGLHV 0HEUDKWX HW DO f FRQFOXGHG WKDW UHVLVWDQFH WR WKH 0H[LFDQ EHDQ EHHWOH ZDV SULPDULO\ DQ H[SUHVVLRQ RI DGGLWLYH JHQHWLF YDULDQFH .LOHQ HW DO f VXJJHVWHG WKDW VXVFHSWLELOLW\ WR WKH VR\EHDQ ORRSHU ZDV SDUWLDOO\ GRPLQDQW DQG D IHZ PDMRU JHQHV FRQGLWLRQHG UHVLVWDQFH %DVHG RQ WKHLU ZRUN ZLWK 3, 3, DQG 3, DQG WKH ) SURJHQ\ GHULYHG IURP WKH LQWHUFURVVHV DPRQJ WKH WKUHH 3LfV /DPEHUW DQG .LOHQ f VXJJHVWHG WKDW WKH JHQHWLF EDVLV IRU LQVHFW UHVLVWDQFH LQ 3, GLIIHUHG IURP WKDW LQ WKH RWKHU WZR 3LfV 7KLV ZDV IXUWKHU VXSSRUWHG E\ WKHLU ZRUN ZLWK WKH WKUHH 3LfV DQG WKH ) OLQHV GHULYHG IURP LQWHUFURVVHV DPRQJ WKH WKUHH 3LfV f 7KH\ K\SRWKHVL]HG WKDW HDFK 3, FRQWDLQHG D GLVWLQFW UHVLVWDQFH JHQH

PAGE 42

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f 7R VWXG\ WKH LQKHULWDQFH RI UHVLVWDQFH WR VR\EHDQ ORRSHU LQ VR\EHDQ WKH EUHHGLQJ OLQH ZDV FURVVHG ZLWK WKH FXOWLYDU %UD[WRQ f KDV ZKLWH IORZHUV Zf JUD\ SXEHVFHQFH f VHQVLWLYLW\ WR PHWULEX]LQ >DPLQROOGLPHWK\OHWK\Of PHWK\OWKLRfOWULD]LQ+fRQH@ KPf UHVLVWDQFH WR 3K\WRSKWKRUD URW LQGXFHG E\ 3K\WRSKWKRUD PHJDVSHUPD 'UHFKV I VS JO\FLQHD 7 .XDQ t '& (UZLQf 5SVOFf DQG UHVLVWDQFH WR VR\EHDQ ORRSHU 7KH GRQRU RI LQVHFW UHVLVWDQFH ZDV WKH JHUPSODVP OLQH f ZKLFK GHULYHG LWV UHVLVWDQFH IURP 3, f 7KH PDOH SDUHQW %UD[WRQ KDV SXUSOH IORZHUV :f WDZQ\ SXEHVFHQFH f WROHUDQFH WR PHWULEX]LQ ^+Pf VXVFHSWLELOLW\ WR 3K\WRSKWKRUD URW USVf DQG VXVFHSWLELOLW\ WR VR\EHDQ ORRSHU %RWK SDUHQWV ZHUH RI PDWXULW\ JURXS 9,, +DQG SROOLQDWLRQV f PDGH LQ WKH ILHOG SURGXFHG VHYHQ VHHGV RI WKH FURVV [ %UD[WRQ

PAGE 43

7KH VHYHQ ) VHHGV SURGXFHG LQ ZHUH VSDFHSODQWHG FP DSDUW LQ FP URZV LQ WKH ) QXUVHU\ 6HYHUDO URZV RI HDFK SDUHQW DQG %UD[WRQ ZHUH SODQWHG DW D UDWH RI VHHGV SHU PHWHU IRU FURVVLQJ SXUSRVHV )RXU ) SODQWV HPHUJHG DQG WKHVH SODQWV ZHUH XVHG DV WKH SROOHQ SDUHQWV IRU WZR EDFNFURVVHV [ )' [ %UD[WRQf %&f DQG %UD[WRQ [ )' [ %UD[WRQf %&f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f1 PHWKR[\DFHW\OfDODQLQH PHWK\O HVWHU@ WR RSWLPL]H UDSLG HPHUJHQFH RI KHDOWK\ SODQWV DQG SODQWHG RQH VHHG SHU SRW LQ / SRWV ILOOHG ZLWK VRLO WUHDWHG ZLWK %UDG\UKL]RELXP MDSRQLFXP LQRFXOXP LQ WKH JUHHQKRXVH GXULQJ WKH ZLQWHU RI 3ODQWV ZHUH DOORZHG WR VHOISROOLQDWH WKXV SURYLGLQJ VHHG RI WKH %6 WKH VHOIHG JHQHUDWLRQ RI %&f DQG %6 WKH VHOIHG JHQHUDWLRQ RI %&f JHQHUDWLRQV $W PDWXULW\ WKH SODQWV ZHUH KDUYHVWHG DQG WKUHVKHG ZLWK DQ $/0$&2 $OOHQ 0DFKLQH &R 1HYDGD ,$ f VLQJOHSODQW WKUHVKHU

PAGE 44

7KH LQKHULWDQFH VWXG\ ZDV FRQGXFWHG LQ D VFUHHQHG ILHOG FDJH P KLJK ; P ZLGH [ P ORQJ f $ VFKHPDWLF RI WKH ILHOG FDJH GHVLJQ LV VKRZQ LQ )LJXUH ,Q PLG$SULO WKH VRLO ZDV GLVNHG WZLFH LQ RSSRVLWH GLUHFWLRQV ZLWK D GLVN KDUURZ 2Q 0D\ WKH VRLO ZDV WLOOHG ZLWK D VSULQJWRRWK KDUURZ WR VPRRWK WKH VXUIDFH 7KLV ZDV IROORZHG ZLWK D SUHSODQWLQFRUSRUDWHG KHUELFLGH WUHDWPHQW RI FORPD]RQH FKORURSKHQ\Of PHWK\OGLPHWK\OLVRR[D]ROLGLQRQH DQG WULIOXUDOLQ >GLQLWUR11 GLSURS\OWULIOXRURPHWK\OfEHQ]HQDPLQH@ DW NJ KD DQG NJ KD UHVSHFWLYHO\ RQ 0D\ 7KH DUHD ZLWKLQ WKH FDJH ZDV PDUNHG ZLWK URZV VSDFHG FP DSDUW WR EH XVHG DV D JXLGH IRU SODQWLQJ ZLWK SXVK SODQWHUV 7KH VHHG IRU WKH VWXG\ ZHUH SODQWHG RQ 0D\ $ VXVFHSWLEOH FXOWLYDU f&HQWHQQLDOf f ZDV SODQWHG DW VHHG SHU PHWHU RI URZ LQ ERUGHU URZV DV ZHOO DV D P VHFWLRQ DW WKH HQGV RI HDFK URZ RI H[SHULPHQWDO PDWHULDO 6HHG WR SURGXFH H[SHULPHQWDO SODQWV ZHUH VFDULILHG WUHDWHG ZLWK PHWDOD[\O DQG VSDFH SODQWHG FP DSDUW 7KH SDUHQWV DQG %UD[WRQ ZHUH SODQWHG LQ WKUHH P URZV IRU D WRWDO RI SODQWV HDFK $OVR VHHG IURP HDFK RI WKH IRXUWHHQ %& DQG VL[ %& SODQWV ZHUH SODQWHG LQ P URZV )ROORZLQJ WKH EDFNFURVV SURJHQ\ ILYH )M VHHG HDFK RI WKH FURVV [ %UD[WRQ DQG LWV UHFLSURFDO ZHUH SODQWHG 7KHVH ZHUH IROORZHG E\ WKH ) VHHG IURP WKH ) [ %UD[WRQf SODQWV JURZQ LQ 7KH LQVHFW UHVLVWDQW JHUPSODVP OLQHV f DQG 3, f ZHUH SODQWHG LQ WKH PLGGOH DQG DW WKH HQG RI WKH H[SHULPHQWDO PDWHULDO IRU D WRWDO RI SODQWV HDFK

PAGE 45

)LJXUH 6FKHPDWLF GUDZLQJ RI ILHOG FDJHV XVHG LQ WKH DQG H[SHULPHQW -

PAGE 46

1RUPDO SURGXFWLRQ SUDFWLFHV ZHUH XVHG WR PDLQWDLQ KHDOWK\ SODQWV 6RLO ZDV FXOWLYDWHG RQ DQG -XQH WR FRQWURO ZHHGV DQG EHQWD]RQ >OPHWK\OHWK\OfO+f OEHQ]RWKLDGLD]LQ+fRQHGLR[LGH@ ZDV DSSOLHG DW NJ KDn RQ -XQH ZLWK D KDQG KHOG SQHXPDWLF VSUD\HU 6XSSOHPHQWDO LUULJDWLRQ ZDV QRW UHTXLUHG GXULQJ WKLV WLPH $ 6DUDQ FDJH WRS f ZDV LQVWDOOHG RQ -XQH DV WKH ILQDO VWHS LQ SUHSDUDWLRQ IRU WKH UHOHDVH RI LQVHFWV $SSUR[LPDWHO\ ODERUDWRU\UHDUHG f VR\EHDQ ORRSHU PRWKV ZHUH UHOHDVHG LQ WKH FDJH DW K RQ -XQH 7KH DEVHQFH RI SK\VLFDO UHVWULFWLRQV PDGH DOO DUHDV RI WKH FDJH HTXDOO\ DFFHVVLEOH WR WKH PRWKV :DWHU ZDV DSSOLHG E\ IXUURZ LUULJDWLRQ WR HOHYDWH WKH UHODWLYH KXPLGLW\ ZLWKLQ WKH FDJH WR HQKDQFH HJJ OD\LQJ $OO SODQWV ZHUH DW WKH 9 WR 9 JURZWK VWDJH DW WKH WLPH RI UHOHDVH /DUYDH HPHUJHG IURP HJJV DIWHU DQ LQFXEDWLRQ SHULRG RI DSSUR[LPDWHO\ WKUHH GD\V /DUYDH ZHUH DOORZHG WR IHHG XQWLO VXVFHSWLEOH &HQWHQQLDO ZDV GHWHUPLQHG WR EH WR b GHIROLDWHG ZKLFK RFFXUUHG RQ -XO\ RU GD\V DIWHU PRWKV ZHUH UHOHDVHG $W WKDW WLPH PHWKRP\O ZDV DSSOLHG DW NJ KD ZLWK D EDFNSDFN VSUD\HU WR NLOO WKH ORRSHUV DQG VWRS IXUWKHU IHHGLQJ 2Q -XO\ WKH 6DUDQ FDJH WRS ZDV UHPRYHG WR IDFLOLWDWH UHJURZWK E\ SODQWV WKDW ZHUH WR EH KDUYHVWHG IRU XVH LQ DQG DOO SODQWV LQ WKH VWXG\ ZHUH HYDOXDWHG IRU GHIROLDWLRQ (DFK SODQW ZDV DVVLJQHG D VFRUH RI OWR $9 PHWKRGf EDVHG RQ WKH YLVXDO HVWLPDWH RI GHIROLDWLRQ 7KH OWR UDWLQJ VFDOH UHSUHVHQWHG b LQFUHPHQWV ZLWK D VFRUH RI UHSUHVHQWLQJ GHIROLDWLRQ RI WR b DQG D VFRUH RI UHSUHVHQWLQJ GHIROLDWLRQ RI WR b

PAGE 47

5DWLQJV IRU WKH WZR SDUHQWV DQG %UD[WRQ DQG WKHLU ) DQG ) SURJHQ\ ZHUH JURXSHG E\ WKHLU GHIROLDWLRQ VFRUHV DQG SORWWHG RQ D JUDSK 7KH GLVWULEXWLRQ RI WKH SDUHQWDO DQG ) SRSXODWLRQV ZHUH WHVWHG IRU QRUPDOLW\ IROORZLQJ WKH SURFHGXUH GHVFULEHG E\ /HRQDUG HW DO f ,QLWLDOO\ WKH GHIROLDWLRQ GDWD RI WKHVH SRSXODWLRQV ZHUH VXPPDUL]HG LQ IUHTXHQF\ GLVWULEXWLRQV E\ XVLQJ WKH XSSHU FODVV OLPLWV 7KH PHDQ [ YDULDQFH V DQG VWDQGDUG GHYLDWLRQ V RI HDFK SRSXODWLRQ ZHUH FDOFXODWHG IURP WKH FODVV FHQWHUV LH f XVLQJ WKH IROORZLQJ IRUPXODV [ QA2f QO f QfQ Q Qf V Q [f QAO [f A [fQ f V Vf ZKHUH Qf ZDV WKH QXPEHU RI SODQWV LQ HDFK RI WKH UHVSHFWLYH FODVVHV DQG Q ZDV WKH JUDQG WRWDO )ROORZLQJ WKH FDOFXODWLRQ RI WKHVH YDULDEOHV WKH YDULDEOH [ ZDV FDOFXODWHG IRU HDFK FODVV LQ WKH IROORZLQJ PDQQHU [ [ XSSHU FODVV OLPLWfV $IWHU GHWHUPLQLQJ WKH [ IRU HDFK FODVV D WDEOH RI QRUPDO SUREDELOLW\ LQWHJUDOV ZDV XVHG WR GHWHUPLQH WKH FRUUHVSRQGLQJ = YDOXH RI HDFK FODVV 7KH WKHRUHWLFDO SHUFHQWDJH RI HDFK SRSXODWLRQ ZDV FDOFXODWHG E\ VXEWUDFWLQJ = IURP DQG PXOWLSO\LQJ E\ 7KH WKHRUHWLFDO QXPEHU RI HDFK FODVV ZDV FDOFXODWHG DV IROORZV Q [ WKHRUHWLFDO b WKHRUHWLFDO Q

PAGE 48

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

PAGE 49

2FWREHU DQG 1RYHPEHU 7KH VRLO LQ WKH FDJH ZDV VXEVRLOHG RQ 1RYHPEHU LQ SUHSDUDWLRQ IRU WKH H[SHULPHQW 7ZR ILHOG FDJHV ZHUH XVHG WR FRPSOHWH WKH LQKHULWDQFH VWXG\ 6RLO LQ HDFK FDJH ZDV GLVNHG ZLWK D GLVN KDUURZ DQG WUHDWHG ZLWK WULIOXUDOLQ DW NJ KD ZKLFK ZDV LQFRUSRUDWHG ZLWK D VSULQJWRRWK KDUURZ RQ $SULO 7R FRQWURO D EURDG VSHFWUXP RI ZHHGV D WDQN PL[ RI FORPD]RQH DW NJ KDn DQG LPD]DTXLQ >GLK\GURPHWK\O PHWK\OHWK\OfR[R +LPLGD]RO\OfTXLQROLQHFDUER[\OLFDFLG@ DW NJ KDn ZDV DSSOLHG DQG LQFRUSRUDWHG ZLWK D VSULQJWRRWK KDUURZ RQ 0D\ 2Q 0D\ VRLO ZDV UHZRUNHG ZLWK D VSULQJWRRWK KDUURZ DQG URZV VSDFHG FP DSDUW ZHUH PDUNHG WR VHUYH DV JXLGHV IRU WKH SXVK SODQWHU (DFK URZ FRQVLVWHG RI SODQWV VSDFHG FP DSDUW H[FHSW IRU WKH )M %& DQG %& JHQHUDWLRQV ZKLFK KDG RQO\ RQH SODQW SHU URZ 7KHUH ZHUH WKUHH URZV FRQVLVWLQJ RI SODQWV HDFK RI WKH LQVHFWUHVLVWDQW JHUPSODVP OLQHV DQG 3, 7KH SODQWLQJ RUGHU IRU HDFK FDJH ZDV JHQHUDWHG E\ UDQGRPL]LQJ WKH URZV ZLWK 3& 6$6 352& 3/$1 f (DFK FDJH FRQWDLQHG WKH HQWLUH FRPSOHPHQW RI SODQWV DQG ZDV WUHDWHG DV D UHSOLFDWLRQ RYHU HQYLURQPHQWV 7KH ERUGHU URZV ZHUH SODQWHG WR &HQWHQQLDO DW VHHG SHU PHWHU LQ HDFK FDJH 7KH URZV RI H[SHULPHQWDO SRSXODWLRQV ZHUH FHQWHUHG FP DSDUW 5DLQIDOO RI PP KDOWHG SODQWLQJ RQ 0D\ LQ FDJH DIWHU RQH WKLUG RI WKH URZV ZHUH SODQWHG 7KH UHPDLQGHU RI FDJH DQG DOO RI FDJH ZDV OLJKWO\ WLOOHG ZLWK D VSULQJWRRWK KDUURZ RQ 0D\ 7KH UHPDLQLQJ URZV LQ FDJH ZHUH SODQWHG RQ 0D\ $SSUR[LPDWHO\ P RI

PAGE 50

&HQWHQQLDO ZDV SODQWHG DW WKH HQGV RI DOO URZV WR VHUYH DV D ERUGHU 7KH H[SHULPHQWDO SRSXODWLRQV ZHUH VXUURXQGHG E\ D VXVFHSWLEOH FXOWLYDU 2Q 0D\ VRLO LQ FDJH ZDV UROOHG ZLWK D ZDWHUILOOHG UROOHU SXOOHG E\ D IRXU ZKHHO DOOWHUUDLQ YHKLFOH 7KLV RSHUDWLRQ VPRRWKHG DQG ILUPHG WKH VHHG EHG ZKLFK GHFUHDVHG WKH UDWH RI HYDSRUDWLRQ RI PRLVWXUH IURP WKH VRLO 7KH URZV ZHUH WKHQ PDUNHG DV LQ FDJH 7KH VXVFHSWLEOH FXOWLYDU &HQWHQQLDO ZDV SODQWHG E\ WKH VDPH PHWKRG DV LQ FDJH 7KH HQWLUH DUHD ZLWKLQ FDJH ZDV SODQWHG RQ 0D\ 6KRUW VHFWLRQV RI WKH &HQWHQQLDO ERUGHU URZV ZHUH UHSODQWHG RQ -XQH 6RLO LQ WKH FDJHV ZDV FXOWLYDWHG RQ DQG -XQH ZLWK D VHOISURSHOOHG JDUGHQ WLOOHU 6XSSOHPHQWDO VRLO PRLVWXUH ZDV SURYLGHG E\ IXUURZ LUULJDWLRQ RQ DQG -XQH WR HQVXUH WKDW WKH SODQWV ZRXOG KDYH VXIILFLHQW YHJHWDWLYH JURZWK DW WKH WLPH RI LQVHFW UHOHDVH 7KH 6DUDQ WRSV ZHUH SXW RQ WKH FDJHV RQ -XQH DV ILQDO SUHSDUDWLRQ IRU WKH UHOHDVH RI LQVHFWV $SSUR[LPDWHO\ ODERUDWRU\UHDUHG f VR\EHDQ ORRSHU PRWKV ZHUH UHOHDVHG LQ HDFK FDJH DW K RQ -XO\ $ UDLQIDOO RI PP RQ -XO\ SURYLGHG VXIILFLHQW VRLO PRLVWXUH WR PDLQWDLQ DQ HOHYDWHG UHODWLYH KXPLGLW\ ZLWKLQ WKH FDJHV GXULQJ WKH HJJOD\LQJ SHULRG $W WKH WLPH RI UHOHDVH WKH SODQWV ZHUH DW WKH 9 WR 9 VWDJH RI JURZWK LQ FDJH DQG 9 WR 9 LQ FDJH f /DUYDH HPHUJHG DIWHU DQ LQFXEDWLRQ SHULRG RI DSSUR[LPDWHO\ WKUHH GD\V /DUYDH ZHUH DOORZHG WR IHHG LQ HDFK FDJH XQWLO -XO\ DW ZKLFK WLPH &HQWHQQLDO ZDV DSSUR[LPDWHO\ b GHIROLDWHG 7KH FDJH WRSV ZHUH WKHQ UHPRYHG WR IDFLOLWDWH WKH GHIROLDWLRQ UDWLQJV DQG WR SURPRWH UHJURZWK RI WKH SODQWV %RWK FDJHV ZHUH VSUD\HG ZLWK

PAGE 51

PHWKRP\O DW NJ KDn WR NLOO ODUYDH 7KLV LQVXUHG WKDW DGGLWLRQDO IHHGLQJ ZRXOG QRW RFFXU RQFH WKH UDWLQJV KDG EHJXQ &DJH ZDV UDWHG RQ DQG -XO\ XVLQJ WKH $9 UDWLQJ PHWKRG 7KLV PHWKRG XVHG WKH VDPH OWR VFDOH SUHYLRXVO\ GHVFULEHG IRU WKH :3 PHWKRG LQ ,Q WKH $9 PHWKRG WKH SODQWV ZHUH YLVXDOO\ SDUWLWLRQHG LQWR WKUHH VHFWLRQV WRS PLGGOH ERWWRPf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fV VFDOLQJ WHVWV f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

PAGE 52

7R GHWHUPLQH LI HDFK RI WKH VFDOLQJ WHVWV KDG WKH H[SHFWHG YDOXH RI ]HUR HDFK ZDV WHVWHG LQ WKH IROORZLQJ PDQQHU W$ $9Dfr K %9Efr WF &9Ffr W' '9Gfr ZKHUH 9$ff$ 9+fn$ 9&fn$ DQG 9'fn$ ZHUH WKH VWDQGDUG HUURUV RI HDFK WHVWV 7KH W YDOXHV REWDLQHG ZHUH FRPSDUHG WR D WDEOH RI W GLVWULEXWLRQ f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f (UURUV DUH LQKHUHQWO\ VPDOOHU VLQFH PHDQV ILUVW RUGHU VWDWLVWLFVf DUH XVHG UDWKHU WKDQ YDULDQFHV VHFRQG RUGHU VWDWLVWLFVf 6PDOOHU H[SHULPHQWV DOORZ WKH VDPH GHJUHH RI SUHFLVLRQ

PAGE 53

$GGLWLYH Df GRPLQDQW Gf DQG HSLVWDWLF DD DG DQG GGf HIIHFWV DUH HVWLPDWHG XVLQJ PHDQV UDWKHU WKDQ YDULDQFHV DQG ,W LV DSSOLFDEOH WR ERWK FURVV DQG VHOISROOLQDWLQJ FURSV ,Q XWLOL]LQJ WKH JHQHUDWLRQ PHDQV DQDO\VLV LW LV DVVXPHG WKDW WKH SDUHQWV DUH KRPR]\JRXV DQG WKHUH LV QR OLQNDJH DPRQJ JHQHV LQIOXHQFLQJ WKH WUDLW EHLQJ VWXGLHG ,I WKLV DVVXPSWLRQ LV FRUUHFW DQG LW DSSHDUV WKDW WKH WUDLW LQ TXHVWLRQ LV QRW TXDQWLWDWLYHO\ LQKHULWHG WKHVH JHQHUDWLRQV FDQ EH XVHG WR DQDO\]H IRU 0HQGHOLDQ UDWLRV +DOODXHU DQG 0LUDQGD f FLWHG WKH IROORZLQJ GLVDGYDQWDJHV WR WKH JHQHUDWLRQ PHDQV DQDO\VLV $Q HVWLPDWH RI KHULWDELOLW\ FDQQRW EH REWDLQHG DQG *HQHWLF DGYDQFHV FDQQRW EH SUHGLFWHG EHFDXVH JHQHWLF YDULDQFHV DUH QRW DYDLODEOH $OWKRXJK WKHVH GLVDGYDQWDJHV DUH LQKHUHQW WR WKH JHQHUDWLRQ PHDQV DQDO\VLV WKH\ DUH RI OLWWOH FRQVHTXHQFH VLQFH DGGLWLRQDO FDOFXODWLRQV FDQ EH PDGH WR REWDLQ WKH YDULDQFHV QHFHVVDU\ WR HVWLPDWH KHULWDELOLW\ *HQHUDWLRQ PHDQV DQDO\VLV DV GHVFULEHG E\ +D\ PDQ f KDV EHHQ XVHG H[WHQVLYHO\ LQ RWKHU FURSV VXFK DV FRP f FRWWRQ *RVV\SLXP KLUVXWXP /f f DQG ZKHDW f ,Q PRVW FDVHV *DPEOHfV f QRWDWLRQ LV XVHG LQ FRQMXQFWLRQ ZLWK +D\PDQfV f PHWKRGRORJ\ *HQHUDWLRQ PHDQV DQDO\VLV DSSDUHQWO\ KDV QRW EHHQ XVHG SUHYLRXVO\ LQ VR\EHDQ WR DQDO\]H TXDQWLWDWLYH WUDLWV %HFDXVH RI WKH DGYDQWDJHV RI

PAGE 54

WKLV DQDO\VLV LW ZDV FKRVHQ WR GHWHUPLQH WKH PRGH RI LQKHULWDQFH RI UHVLVWDQFH WR VR\EHDQ ORRSHU 3RSXODWLRQV XVHG ZHUH GHULYHG IURP WKH FURVV [ %UD[WRQ )LJXUH f $ PLQLPXP RI VL[ SRSXODWLRQV DUH UHTXLUHG LQ D JHQHUDWLRQ PHDQV DQDO\VLV ZLWK D VL[ SDUDPHWHU PRGHO EXW WKLV H[SHULPHQW ZDV H[SDQGHG WR LQFOXGH QLQH SRSXODWLRQV 7KH QXPEHU RI LQGLYLGXDOV FRPSULVLQJ HDFK JHQHUDWLRQ DUH OLVWHG LQ 7DEOH 7R GHWHUPLQH WKH LQKHULWDQFH RI UHVLVWDQFH WR VR\EHDQ ORRSHU LQ VR\EHDQ DFFRUGLQJ WR +D\PDQfV f PRGHO SURJUDPV ZHUH ZULWWHQ LQ 6$6 WR DQDO\]H WKH GDWD E\ FDJH VHH $SSHQGL[ &f DQG FRPELQHG RYHU FDJHV ZKHUH FDJHV ZHUH WUHDWHG DV EORFNV VHH $SSHQGL[ 'f 8VLQJ +D\PDQfV PHWKRGRORJ\ DQG *DPEOHfV f QRWDWLRQ WKH PRGHOV IRU WKH JHQHUDWLRQ PHDQV DQDO\VLV ZHUH DV IROORZV S P D G DD DG GG S f§ P D G DD DG GG ) P G GG I P %& P D DD EF P D DD I P G GG %6M P D G DD DG GG EV P D G DD DG GG ZKHUH P WKH RYHUDOO PHDQ DQG D G DD DG DQG GG UHSUHVHQW WKH DGGLWLYH GRPLQDQFH DGGLWLYH [ DGGLWLYH DGGLWLYH [ GRPLQDQFH DQG GRPLQDQFH [ GRPLQDQFH JHQHWLF HIIHFWV UHVSHFWLYHO\ (DFK 6$6 SURJUDP ILWWHG WZR UHJUHVVLRQ PRGHOV ZKLFK ZHUH VHW XS XVLQJ PDWUL[ QRWDWLRQ DFFRUGLQJ WR WKH SURFHGXUHV RXWOLQHG E\ -HQQLQJV HW DO f 7KH ILUVW

PAGE 55

' 3f [ %UD[WRQ 3A 3 [ ) [ 3 %& %& %6 %6 )LJXUH 0DWLQJ VFKHPH WR GHULYH SRSXODWLRQV UHTXLUHG WR HVWLPDWH WKH JHQHWLF HIIHFWV RI UHVLVWDQFH WR WKH VR\EHDQ ORRSHU LQ VR\EHDQ

PAGE 56

7DEOH 3RSXODWLRQV XVHG IRU JHQHUDWLRQ PHDQV DQDO\VLV RI WKH LQKHULWDQFH RI UHVLVWDQFH WR WKH VR\EHDQ ORRSHU 3RSXODWLRQ *HQ 5RZ 1XPEHU 3ODQWV SHU 5RZI 7RWDO 3ODQWV 3 %UD[WRQ 3 [ %UD[WRQ ) ; %UD[WRQ I [ %UD[WRQ I 'f ; %UD[WRQ %& 'f ; %UD[WRQ %6 %UD[WRQf [ EF %UD[WRQf [ EV 3, I5RZV ZHUH FP ORQJ H[FHSW ZKHUH WKHUH ZDV RQO\ SODQWURZ A,QVHFWUHVLVWDQW JHUPSODVP OLQHV XVHG DV FKHFNV

PAGE 57

UHJUHVVLRQ PRGHO GHILQHG DV 0RGHO FRQVLVWHG RI WKH WKUHH SDUDPHWHUV P D DQG G 7KH VHFRQG UHJUHVVLRQ PRGHO GHILQHG DV 0RGHO FRQVLVWHG RI WKH HSLVWDWLF HIIHFWV DD DG DQG GG LQ DGGLWLRQ WR WKH SDUDPHWHUV LQ 0RGHO 0RGHO LV XVHG RQO\ LI D VLJQLILFDQW DGGLWLYH RU GRPLQDQW HIIHFW LV GHWHFWHG DQG WR GHWHUPLQH LI VLJQLILFDQW HSLVWDWLF HIIHFWV H[LVW WKDW DUH FRQWULEXWLQJ WR WKH VLJQLILFDQFH LQ 0RGHO 7R DGMXVW IRU WKH XQHTXDO SRSXODWLRQ VL]HV FRPSULVLQJ HDFK JHQHUDWLRQ WKH PRGHOV ZHUH ZHLJKWHG XVLQJ UHFLSURFDOV RI WKH VWDQGDUG HUURUV RI WKH JHQHUDWLRQ PHDQV DV VXJJHVWHG E\ 5RZH DQG $OH[DQGHU f ,Q DGGLWLRQ 3RZHUVf SDUWLWLRQLQJ PHWKRG RI JHQHWLF DQDO\VLV f ZDV LQFRUSRUDWHG WR WHVW SURSRVHG JHQHWLF PRGHOV ,QLWLDOO\ WKH IUHTXHQF\ GLVWULEXWLRQV RI WKH 3O 3 ) DQG ) JHQHUDWLRQV ZHUH WHVWHG IRU QRUPDOLW\ E\ WHVWLQJ WKH REVHUYHG YHUVXV WKHRUHWLFDO GLVWULEXWLRQ ZLWK D ; WHVW IROORZLQJ WKH SURFHGXUH SUHYLRXVO\ RXWOLQHG )UHTXHQF\ GLVWULEXWLRQV REVHUYHG DQG WKHRUHWLFDOf RI WKH ) SRSXODWLRQ ZHUH FRQYHUWHG WR SHUFHQWDJHV LQ HDFK FODVV LQ WKH IROORZLQJ PDQQHU bLQ HDFK FODVVf QcQ ZKHUH Qc ZDV WKH QXPEHU RI SODQWV LQ HDFK FODVV DQG Q ZDV WKH WRWDO QXPEHU RI SODQWV LQ WKH ) SRSXODWLRQ $IWHU FRQYHUVLRQ WKH REVHUYHG GLVWULEXWLRQ RI WKH ) SRSXODWLRQ ZDV H[DPLQHG IRU PRGHV )LQDOO\ WKH ) SRSXODWLRQ RI HDFK FDJH ZDV WHVWHG IRU JRRGQHVV RI ILW WR D WZRJHQH DGGLWLYH PRGHO 7KH SURSRVHG JHQHWLF UDWLR IRU WKH PRGHO ZDV DV IROORZV 7R WHVW WKH PRGHO WKH UDWLR ZDV VLPSOLILHG WR E\ FRPELQLQJ WKH GHIROLDWLRQ FODVVHV LQ WKH WZR FDJHV DV IROORZV DQG LQ

PAGE 58

FDJH DQG DQG LQ FDJH 7KH HQG FODVVHV ZHUH GHILQHG E\ D GLVWLQFW WUDQVLWLRQ EHWZHHQ GHIROLDWLRQ FODVV LQ WKH GLVWULEXWLRQ RI WKH ) SRSXODWLRQ RI HDFK FDJH 7KH KRPR]\JRXV SRSXODWLRQV %UD[WRQ DQG 3, ZHUH XVHG WR GHWHUPLQH LI WKHUH ZDV D VLJQLILFDQW ORFDWLRQ HIIHFW ZLWKLQ HDFK FDJH $Q DQDO\VLV RI YDULDQFH ZDV SHUIRUPHG RQ HDFK SRSXODWLRQ WR REWDLQ WKH URZ DQG SODQW ZLWKLQ URZ PHDQ VTXDUHV (VWLPDWHV IRU SODQW WR SODQW YDULDELOLW\ ZLWKLQ D URZ DSf DQG WKH DGGLWLRQDO YDULDELOLW\ RI SODQWV LQ GLIIHUHQW URZV Rf ZHUH REWDLQHG IURP WKH H[SHFWHG YDOXHV RI WKH PHDQ VTXDUHV $OVR )WHVWV IRU VLJQLILFDQW URZ HIIHFWV +RRU f ZHUH SHUIRUPHG RQ HDFK RI WKH KRPR]\JRXV JHQHUDWLRQV ZLWKLQ HDFK FDJH $Q HVWLPDWH IRU WKH QXPEHU RI JHQHV Qf LQYROYHG LQ UHVLVWDQFH WR VR\EHDQ ORRSHU ZDV REWDLQHG IRU WKH SURJHQ\ LQ HDFK FDJH E\ WKH IRUPXOD Q [3 [3f>R)f YD3 YM3OY-D)fYOYYLf@ ,Q WKH IRUPXOD [3M DQG [3 ZHUH WKH PHDQ GHIROLDWLRQ UDWLQJV IRU WKH SDUHQWV DQG A3M R3 DQG MA) ZHUH WKH YDULDQFHV RI WKH UHVSHFWLYH JHQHUDWLRQV 7KLV IRUPXOD LV D PRGLILFDWLRQ RI WKH IRUPXOD SUHVHQWHG E\ 3RHKOPDQ f ,Q WKH PRGLILFDWLRQ D) LV UHSODFHG ZLWK YD3 YD3YLD)fYL YYLf 7KLV PRGLILFDWLRQ ZDV XVHG LQ RUGHU WR EHWWHU HVWLPDWH WKH HQYLURQPHQWDO YDULDQFH VLQFH D)E RA DQG R3 ZHUH HVVHQWLDOO\ HVWLPDWHV RI HQYLURQPHQWDO YDULDWLRQ 7KH GHJUHHV RI IUHHGRP Y Y DQG Yf IRU WKH DVVRFLDWHG JHQHUDWLRQV ZHUH HPSOR\HG WR OHVVHQ SRSXODWLRQ VL]H HIIHFWV RQ WKH HVWLPDWH RI HQYLURQPHQWDO YDULDQFH 7KLV PHWKRG RI HVWLPDWLQJ JHQHV DVVXPHV WKDW WKH JHQHV KDYH HTXDO HIIHFWV ZLWKRXW VLJQLILFDQW GRPLQDQW RU HSLVWDWLF HIIHFWV

PAGE 59

+HULWDELOLW\ RI LQVHFW UHVLVWDQFH LQ WKH EURDG VHQVH ZDV HVWLPDWHG IRU WKH SURJHQ\ LQ HDFK FDJH IURP WKH IRUPXOD + 9R9Sf [ b ZKHUH 9* ZDV WKH JHQHWLF YDULDQFH DQG 93 ZDV WKH SKHQRW\SLF YDULDQFH 7KH JHQHWLF YDULDQFH ZDV 9* 93 9( ZKHUH 9( ZDV WKH HQYLURQPHQWDO YDULDQFH 7KH 9( ZDV HVWLPDWHG E\ REWDLQLQJ WKH ZHLJKWHG DYHUDJH RI WKH YDULDQFHV RI WKH ) SRSXODWLRQ DQG WKH SRSXODWLRQV RI WKH SDUHQWV 3 DQG 3 Y F)3 YF)39MF))A2YYMf ,Q WKH IRUPXOD WKH UHVSHFWLYH GHJUHHV RI IUHHGRP Y ZHUH XVHG WR ZHLJKW WKH YDULDQFHV RI HDFK SRSXODWLRQ 7KH 93 ZDV HVWLPDWHG E\ XWLOL]LQJ WKH YDULDQFH RI WKH ) SRSXODWLRQ 5HVXOWV DQG 'LVFXVVLRQ 7KH GHIROLDWLRQ GDWD RI WKH 3 3 ) DQG ) SRSXODWLRQV IURP WKH SUHOLPLQDU\ FDJH VWXG\ DUH SUHVHQWHG LQ 7DEOH $OVR SUHVHQWHG LQ WKLV WDEOH DUH WKH GHIROLDWLRQ UDWLQJV RI 3, DQG IURP ZKLFK GHULYHG LWV UHVLVWDQFH WR VR\EHDQ ORRSHU ZDV YLVLEO\ PRUH UHVLVWDQW WKDQ WKH RULJLQDO VRXUFH RI

PAGE 60

7DEOH 5DWLQJV RI OHDI IHHGLQJ E\ VR\EHDQ ORRSHU RQ VR\EHDQ SDUHQWV WKHLU ) DQG ) SURJHQ\ DQG WKH JHUPSODVP OLQHV 3, DQG LQ WKH ILHOG FDJH 8SSHU FODVV OLPLWV RI OHDI IHHGLQJ UDWLQJV Q ; '3f %UD[WRQ3f 3 ; 3)Mf 3 ; 3A)Mf 3L ; 3)f 3, WO WR b GHIROLDWLRQ WR b WR b WR b WR b WR b WR b WR b WR b WR b 1RQH RI WKH SODQWV UHFHLYHG D UDWLQJ RI

PAGE 61

UHVLVWDQFH 3, 7KLV VXJJHVWV WKDW DGGLWLRQDO JHQHV IRU UHVLVWDQFH PLJKW EH FRQWULEXWLQJ WR WKH LQFUHDVHG OHYHO RI UHVLVWDQFH 5HVLVWDQFH WR VR\EHDQ ORRSHU ZDV LQWHUSUHWHG DV EHLQJ TXDQWLWDWLYHO\ LQKHULWHG EDVHG RQ WKH GHIROLDWLRQ GDWD RI WKH 3" 3 ) DQG ) SRSXODWLRQV 7DEOH f 'LVWULEXWLRQ RI WKH ) SRSXODWLRQ Q f ZDV FRQWLQXRXV ZLWK D PHDQ GHIROLDWLRQ UDWLQJ RI ZKLFK ZDV VOLJKWO\ OHVV WKDQ WKH SDUHQWDO PLGSRLQW RI $OWKRXJK D KLJK QXPEHU RI LQGLYLGXDOV IHOO LQ WKH IRXUWK UDWLQJ FODVV WR b GHIROLDWLRQf WKH ) SRSXODWLRQ DSSHDUHG WR KDYH D QRUPDO GLVWULEXWLRQ 7KH 3 3 DQG ) SRSXODWLRQ GLVWULEXWLRQV ZHUH WHVWHG IRU QRUPDOLW\ XVLQJ WKH SURFHGXUH RXWOLQHG E\ /HRQDUG HW DO f VHH $SSHQGL[ (f %RWK SDUHQWDO SRSXODWLRQV KDG ORZ ; YDOXHV 7DEOH f LQGLFDWLQJ WKH\ ZHUH QRUPDOO\ GLVWULEXWHG 7KH IUHTXHQF\ GLVWULEXWLRQ RI WKH ) UDWLQJV IDLOHG WKH QRUPDO GLVWULEXWLRQ WHVW 7DEOH f 5DWLQJ FODVV KDG WRR PDQ\ SODQWV ZKHUHDV FODVVHV DQG KDG WRR IHZ 7DEOH DQG $SSHQGL[ (f $OWKRXJK WKH GHIROLDWLRQ UDWLQJV RI WKH ) SRSXODWLRQ GLG QRW ILW D QRUPDO FXUYH WKH GLVWULEXWLRQ ZDV FRQWLQXRXV ZLWK RQO\ RQH PRGH 7KHUHIRUH XVH RI D PHWKRG RI GDWD DQDO\VLV IRU TXDQWLWDWLYH WUDLWV VHHPHG DSSURSULDWH LQ RUGHU WR GHILQH WKH PRGH RI LQKHULWDQFH RI VR\EHDQ ORRSHU UHVLVWDQFH $ JHQHUDWLRQ PHDQV DQDO\VLV f ZDV FKRVHQ DQG DGGLWLRQDO FURVVHV ZHUH PDGH WR VXSSO\ WKH JHQHUDWLRQV QHFHVVDU\ IRU WKH H[SHULPHQW /DUYDH ZHUH DOORZHG WR IHHG XQWLO VXVFHSWLEOH &HQWHQQLDO ZDV b GHIROLDWHG 3ODQWV ZHUH IRXQG LQ DOO UDWLQJ FODVVHV H[FHSW FODVV RQH $OO SRSXODWLRQV UHFHLYLQJ ODUJHU

PAGE 62

7DEOH *RRGQHVV RI ILW WHVW IRU QRUPDOLW\I RI VR\EHDQ ORRSHU GHIROLDWLRQ IRU SRSXODWLRQV RI %UD[WRQ DQG [ %UD[WRQ ) SRSXODWLRQV LQ WKH FDJH VWXG\ 3RSXODWLRQ 'HJUHHV RI )UHHGRP ; 3 %UD[WRQ I I 7HVW IROORZV WKH SURFHGXUH RI /HRQDUG HW DO f

PAGE 63

PHDQ GHIROLDWLRQ VFRUHV LQ 7DEOH f WKDQ LQ 7DEOH f ZKHQ ODUYDH ZHUH NLOOHG ZKHQ &HQWHQQLDO ZDV RQO\ b GHIROLDWHG ,Q WKH $9 UDWLQJ PHWKRG ZDV VXEVWLWXWHG IRU WKH :3 PHWKRG XVHG LQ 7KH $9 PHWKRG DFFRXQWHG IRU IHHGLQJ SUHIHUHQFHV RI VR\EHDQ ORRSHU f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f 6LQFH WKH ) GDWD GLG QRW ILW H[SHFWHG GLVWULEXWLRQV LW ZDV VXVSHFWHG WKDW WKHUH ZDV D ORFDWLRQ HIIHFW ZLWKLQ HDFK FDJH LQWHUIHULQJ ZLWK WKH GHIROLDWLRQ UDWLQJV 7KH IRXU KRPR]\JRXV SRSXODWLRQV %UD[WRQ DQG 3, f LQFOXGHG LQ WKH VWXG\ ZHUH WHVWHG WR GHWHUPLQH LI WKHUH ZDV D VLJQLILFDQW ORFDWLRQ HIIHFW ZLWKLQ HDFK FDJH $ VLJQLILFDQW ORFDWLRQ HIIHFW ZDV GHWHFWHG IRU WKH SRSXODWLRQV RI %UD[WRQ DQG LQ HDFK FDJH 7DEOH f

PAGE 64

7DEOH 5DWLQJV RI OHDI IHHGLQJ E\ VR\EHDQ ORRSHU RQ VR\EHDQ SDUHQWV WKHLU ) )f ) %& %6O %& DQG %6 SURJHQ\ DQG WKH JHUPSODVP OLQHV 3, DQG LQ WKH ILHOG FDJHV 8SSHU &ODVV /LPLWV RI /HDI IHHGLQJ UDWLQJV Q [ &DJH '3f %UD[WRQ3f 3L ; 3)f 3 ; 3)f 3 ; 3)f 3 [ 3 [ 3f%&f 3 [ 3 [ 3f%6f 3 [ 3 [ 3f%&f 3 [ 3 [ 3f%6f 3, &DJH '3f %UD[WRQ3f 3 ; 3)f 3 ; 3)f 3 ; 3)f 3 [ 3 [ 3A%4f 3 [ 3 [ 3f%6f 3 [ 3L [ 3f%&f 3 [ 3L [ 3f%6f 3, I%DVHG RQ WKH PHDQ RI WKUHH REVHUYDWLRQV SHU SODQW WR b GHIROLDWLRQ WR b WR b WR b WR b WR b WR b WR b WR b WR b 1RQH RI WKH SODQWV UHFHLYHG D UDWLQJ RI A'HULYHG IURP WKH PHDQ GHIROLDWLRQ RI DOO SODQWV ZLWKRXW UHJDUG RI WKH XSSHU FODVV OLPLWV

PAGE 65

7DEOH (VWLPDWHV RI YDULDQFH FRPSRQHQWV REWDLQHG IURP DQ DQDO\VLV RI WKH KRPRJHQHRXV SRSXODWLRQV %UD[WRQ 3, IRU D ORFDWLRQ HIIHFW GXH WR IHHGLQJ E\ VR\EHDQ ORRSHU LQ ILHOG FDJHV LQ 3RSXODWLRQ FULS R9 D fQ ,, &DJH 3f rrr %UD[WRQ 3f rrr %'f fr 3, 'f r &DJH 3f rrr %UD[WRQ 3f r %'f rf 3, 'f r rrf 6LJQLILFDQW DW WKH DQG SUREDELOLW\ OHYHOV UHVSHFWLYHO\ I %DVHG RQ GHIROLDWLRQ UDWLQJV PDGH ZLWK D WR VFDOH GLYLGHG LQWR b LQFUHPHQWV ZLWK D UHIOHFWLQJ b GDPDJH r 9DULDQFH RI SODQWV ZLWKLQ VDPH URZ i 9DULDQFH RI SODQWV DPRQJ URZV

PAGE 66

7KH ORFDWLRQ HIIHFW IRU WKH 3, SRSXODWLRQ ZDV VLJQLILFDQW S 7DEOH f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f DUH DOVR SUHVHQWHG ,Q DOO LQVWDQFHV H[FHSW WKH %& JHQHUDWLRQ LQ FDJH WKH SUHGLFWHG PHDQV ZHUH ZLWKLQ s RQH VWDQGDUG GHYLDWLRQ RI WKH REVHUYHG PHDQV LQ HDFK FDJH 7KH REVHUYHG PHDQ IRU WKH %& JHQHUDWLRQ LQ HDFK FDJH ZDV ODUJHU WKDQ WKH SUHGLFWHG PHDQ 7KLV GHYLDWLRQ IURP WKH SUHGLFWHG PHDQ FRXOG EH DVVRFLDWHG ZLWK WKH VPDOO VDPSOH VL]H DYDLODEOH IRU GHIROLDWLRQ UDWLQJV 9DULDQFHV RI WKH JHQHUDWLRQV IURP HDFK FDJH DUH SUHVHQWHG LQ 7DEOH DQG DUH XVHG ODWHU LQ 0DWKHUfV VFDOLQJ WHVW f DQG WR GHWHUPLQH KHULWDELOLW\ 7KH UHVXOWV RI 0DWKHUfV VFDOLQJ WHVW f IRU WKH FURVV [ %UD[WRQ DUH VKRZQ LQ 7DEOH $FFRUGLQJ WR 0DWKHU DQG -LQNV f LI WKH YDOXHV RI $ % & DQG GR QRW GLIIHU IURP ]HUR WKHQ DGGLWLYH JHQH HIIHFWV DUH LQGLFDWHG 7KH VLJQLILFDQFH RI DQ\ RI WKHVH VFDOHV LQGLFDWHV WKH SUHVHQFH RI HSLVWDVLV ,Q FDJH QRQVLJQLILFDQW YDOXHV IRU $ % & DQG LQGLFDWHG WKDW DQ DGGLWLYHGRPLQDQFH PRGHO ZDV DGHTXDWH &DJH GDWD \LHOGHG VLPLODU UWHVWV IRU $ & DQG VXSSRUWLQJ WKH FRQFOXVLRQV IURP FDJH 7KH % VFDOH WHVW FRXOG QRW EH DSSOLHG WR FDJH VLQFH WKH %& SRSXODWLRQ KDG D YDULDQFH

PAGE 67

7DEOH 0HDQV VWDQGDUG HUURUV DQG YDULDQFHV RI VR\EHDQ ORRSHU GHIROLDWLRQ UDWLQJV IRU SDUHQWDO OLQHV DQG VHYHQ SRSXODWLRQV RI VR\EHDQ DULVLQJ IURP WKH FURVV [ %UD[WRQ JURZQ LQ ILHOG FDJHV DW 6WRQHYLOOH 06 LQ 3RSXODWLRQ *HQ 1XPEHU $FWXDO 0HDQ 3UHGI IIILn 6( D 3 &DJH %UD[WRQ 3 3L ; 3 ) 3L [ 3 I 3L [ 3 I 3L [ 3L [ 3f %& 3L r 3L [ 3f %6 3 [ 3L [ 3f EF 3 ; 3 ; 3f %6 3L &DJH %UD[WRQ 3 3L [ 3 ) 3L ; 3 I 3L [ 3 I 3L r 3L r 3f %& 3L [ 3L [ 3f %6 3 ; 3M [ 3f EF 3 ; 3M [ 3f EV fcf%DVHG RQ UHODWLRQVKLSV GHVFULEHG E\ 0DWKHU DQG -LQNV A6WDQGDUG HUURU

PAGE 68

7DEOH 0DWKHUfV VFDOLQJ WHVW DSSOLHG WR WKH FURVV [ %UD[WRQ WR WHVW DGHTXDF\ RI DGGLWLYHGRPLQDQFH PRGHO IRU UHVLVWDQFH WR VR\EHDQ ORRSHU 7HVW &DJH &DJH $

PAGE 69

RI ]HUR 6LQFH & ZDV QRW VLJQLILFDQWO\ GLIIHUHQW IURP ]HUR LQ HLWKHU FDJH 7DEOH f WKLV LQGLFDWHG WKDW JHQHUDWLRQ PHDQV GHSHQGHG RQO\ RQ WKH DGGLWLYH JHQH HIIHFWV DQG QR HSLVWDVLV ZDV SUHVHQW 7KH UHVXOWV RI WKH JHQHUDWLRQ PHDQV DQDO\VLV RI HDFK FDJH DUH UHSRUWHG LQ 7DEOH 7DEOH UHSRUWV WKH UHVXOWV RI WKH FRPELQHG DQDO\VLV ZKHUH WKH FDJH HIIHFWV DUH WUHDWHG DV EORFNV 6LJQLILFDQW QHJDWLYH DGGLWLYH HIIHFWV ZHUH GHWHFWHG ZLWK 0RGHO ZKLFK PHDVXUHV RQO\ DGGLWLYH DQG GRPLQDQW HIIHFWV IRU ERWK &DJHV DQG S DQG S UHVSHFWLYHO\ 7DEOH f DQG WKH FRPELQHG FDJHV S 7DEOH f 7KH VLJQ RI WKLV HIIHFW LV GHSHQGHQW XSRQ WKH SDUHQW GHVLJQDWHG DV 3M 7KH VLJQ RI WKH HIIHFW LV D IXQFWLRQ RI WKH IDFW WKDW DQ LQYHUWHG VFDOH ZDV XVHG DQG DOVR D UHIOHFWLRQ RI WKH UHODWLRQVKLS EHWZHHQ WKH PLGSDUHQW DQG WKH PHDQV RI WKH ) ) DQG ) JHQHUDWLRQV LQGLFDWLQJ ZKLFK SDUHQW ZDV FRQWULEXWLQJ WR WKH DGGLWLYH YDULDWLRQ f ,Q &DJH 7DEOH f WKH PHDQV RI WKH ) ) DQG ) JHQHUDWLRQV ZHUH EHWZHHQ WKH PLGSDUHQW f DQG 3 7KH PHDQV RI WKH ) DQG ) LQ &DJH ZHUH VNHZHG DZD\ IURP WKH PLGSDUHQW f LQ WKH VDPH GLUHFWLRQ EXW WKH ) PHDQ ZDV RULHQWHG WRZDUGV WKH RWKHU SDUHQW $OWKRXJK WKH )MfV PHDQ ZDV RQ WKH RSSRVLWH VLGH LW ZDV ZLWKLQ RQH VWDQGDUG GHYLDWLRQ RI WKH PHDQ $OO RI WKH SURJHQ\ PHDQV VNHZHG WRZDUGV 3 LQ DGGLWLRQ WR D VLJQLILFDQW QHJDWLYH DGGLWLYH HIIHFW LQGLFDWHG WKDW %UD[WRQ PLJKW EH FRQWULEXWLQJ WR WKH DGGLWLYH YDULDWLRQ 6LQFH LW KDV EHHQ UHSRUWHG WKDW %UD[WRQ GRHV KDYH VRPH OHYHO RI UHVLVWDQFH f LW VWDQGV WR UHDVRQ WKDW LW ZRXOG FRQWULEXWH WR WKH DGGLWLYH HIIHFW 7KH IXOO PRGHO 0RGHO ZDV WKHQ ILWWHG WR HVWLPDWH WKH HSLVWDWLF HIIHFWV DV ZHOO DV WKH UHPDLQLQJ DGGLWLYH DQG GRPLQDQW HIIHFWV 1RQH RI WKH HSLVWDWLF DD DG DQG GGf

PAGE 70

7DEOH (VWLPDWHV RI WKH DGGLWLYH GRPLQDQW DQG HSLVWDWLF HIIHFWV LQ WKH JHQHUDWLRQ PHDQV IRU GHIROLDWLRQ E\ VR\EHDQ ORRSHU LQ WKH QLQH SRSXODWLRQV RI WKH [ %UD[WRQ PDWHULDO JURZQ LQ WKH ILHOG FDJHV DW 6WRQHYLOOH 06 LQ (IIHFWV (VWLPDWH 6WDQGDUG HUURU W YDOXH &DJH 0RGHO ) PHDQ Pf rrr $GGLWLYH Df W 'RPLQDQFH Gf 0RGHO ) PHDQ Pf rr $GGLWLYH Df 'RPLQDQFH Gf $GGLWLYH [ $GGLWLYH DDf $GGLWLYH [ 'RPLQDQFH DGf 'RPLQDQFH [ 'RPLQDQFH GGf &DJH 0RGHO ) PHDQ Pf rrr $GGLWLYH Df r 'RPLQDQFH Gf 0RGHO ) PHDQ Pf rr $GGLWLYH Df 'RPLQDQFH Gf $GGLWLYH [ $GGLWLYH DGf $GGLWLYH [ 'RPLQDQFH DGf 'RPLQDQFH ; 'RPLQDQFH GGf rrrfr 6LJQLILFDQW DW WKH DQG SUREDELOLW\ OHYHOV UHVSHFWLYHO\ W 6LJQLILFDQW DW WKH SUREDELOLW\ OHYHO

PAGE 71

7DEOH (VWLPDWHV RI WKH DGGLWLYH GRPLQDQW DQG HSLVWDWLF HIIHFWV LQ WKH JHQHUDWLRQ PHDQV IRU GHIROLDWLRQ E\ VR\EHDQ ORRSHU LQ WKH QLQH SRSXODWLRQV RI WKH [ %UD[WRQ PDWHULDO JURZQ LQ WKH ILHOG FDJHV DW 6WRQHYLOOH 06 LQ (IIHFWV (VWLPDWH 6WDQGDUG HUURU W YDOXH &DHHV DQG &RPELQHG 0RGHO ) PHDQ Pf rrr $GGLWLYH Df rrr 'RPLQDQFH Gf 0RGHO ) PHDQ Pf rrr $GGLWLYH Df 'RPLQDQFH Gf $GGLWLYH ; $GGLWLYH ^DGf $GGLWLYH [ 'RPLQDQFH DGf r 'RPLQDQFH [ 'RPLQDQFH GGf rrr 6LJQLILFDQW DW WKH DQG SUREDELOLW\ OHYHOVUHVSHFWLYHO\

PAGE 72

DGGLWLYH RU GRPLQDQW HIIHFWV ZHUH VLJQLILFDQWO\ GLIIHUHQW IURP ]HUR LQ HLWKHU FDJH $OWKRXJK XQH[SHFWHG DG ZDV VLJQLILFDQW S f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f 3RZHUfV SDUWLWLRQLQJ DQDO\VLV ZDV HPSOR\HG WR DQDO\]H WKH GDWD DQG WHVW WKH SUHGLFWHG JHQH PRGHOV IROORZLQJ WKH H[DPSOHV RI 'HUHQ DQG 4XHVHQEHUU\ f DQG 6DJH DQG GH ,VWXUL] f ,QLWLDOO\ WKH IUHTXHQF\ GLVWULEXWLRQV RI WKH GHIROLDWLRQ GDWD IURP WKH SDUHQWDO )O DQG ) SRSXODWLRQV IURP HDFK FDJH ZHUH WHVWHG IRU QRUPDOLW\ 7DEOH f $V ZDV VHHQ LQ ERWK SDUHQWDO SRSXODWLRQV KDG ORZ ; YDOXHV ZKLFK LQGLFDWHG WKDW WKHVH SRSXODWLRQV ZHUH QRUPDOO\ GLVWULEXWHG DERXW D FHQWUDO PHDQ ,Q FDJH WKH ) SRSXODWLRQ GLG QRW KDYH D QRUPDO GLVWULEXWLRQ EXW WKH ) SRSXODWLRQfV GLVWULEXWLRQ GLG ILW WKH QRUPDO FXUYH 7KH RSSRVLWH VLWXDWLRQ RFFXUUHG ZLWK WKH ) DQG ) SRSXODWLRQV LQ FDJH 7KH GLIIHUHQFHV EHWZHHQ WKH WZR FDJHV ZLWK UHVSHFW WR WKH ) DQG ) SRSXODWLRQV PD\ EH GXH WR D ORFDWLRQ HIIHFW SUHVHQW ZLWKLQ WKH FDJHV ,Q WKH K\SRWKHVL]HG WZRJHQH DGGLWLYH PRGHO WKH PLGGOH FODVVHV ZHUH FRPELQHG UHVXOWLQJ LQ WKUHH FODVVHV IRU WKH PRGHO WR EH WHVWHG 3DUWLWLRQLQJ WKH PLGGOH FODVVHV

PAGE 73

7DEOH *RRGQHVV RI ILW WHVW IRU QRUPDOLW\ RI VR\EHDQ ORRSHU GHIROLDWLRQ GDWD IURP FDJH VWXG\ 3RSXODWLRQ 'HJUHHV RI )UHHGRP ; 3 &DJH %UD[WRQ ) I &DJH %UD[WRQ ) I !

PAGE 74

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fV SDUWLWLRQLQJ DQDO\VLV SUREDEO\ KDV QR XWLOLW\ LQ WKHVH DQDO\VHV $Q HVWLPDWH IRU WKH QXPEHU RI JHQHV Q LQ WKLV FURVV FRQWULEXWLQJ WR WKH H[SUHVVLRQ RI UHVLVWDQFH WR VR\EHDQ ORRSHU ZDV REWDLQHG IRU ERWK FDJHV XVLQJ D PRGLILFDWLRQ RI WKH IRUPXOD SUHVHQWHG E\ 3RHKOPDQ f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

PAGE 75

VHQVH KHULWDELOLWLHV 6LQFH WKH )M JHQHUDWLRQV LQ ERWK FDJHV H[KLELWHG D ELPRGDO GLVWULEXWLRQ IRU GHIROLDWLRQ VFRUHV WKH HQYLURQPHQWDO YDULDQFH ZDV HVWLPDWHG E\ WKH IROORZLQJ IRUPXOD 9S3 9S 3 9S) f9S Y3L Y3Of DV ZDV GRQH LQ HVWLPDWLQJ WKH QXPEHU RI JHQHV 7KLV JLYHV D EHWWHU HVWLPDWH RI WKH HQYLURQPHQWDO YDULDQFH WKDQ MXVW XVLQJ WKH FA) EHFDXVH WKH QXPEHU RI LQGLYLGXDOV FRPSULVLQJ WKH SDUHQWDO JHQHUDWLRQV ZHUH UHODWLYHO\ ODUJH 7DEOH f $OVR SDUHQWV DUH JHQHWLFDOO\ XQLIRUP DQG VKRXOG RQO\ H[KLELW YDULDELOLW\ GXH WR WKH HQYLURQPHQW 7KH KHULWDELOLWLHV ZHUH HVWLPDWHG DW b IRU FDJH DQG b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b GHIROLDWLRQ ZKHUHDV &HQWHQQLDO ZDV DOPRVW FRPSOHWHO\ GHIROLDWHG :KHQ FRPSDUHG WR &HQWHQQLDO ZKLFK LV NQRZQ IRU LWV KLJK VXVFHSWLELOLW\ f %UD[WRQ KDV UHSHDWHGO\ GHPRQVWUDWHG WR b OHVV GHIROLDWLRQ 6LQFH WKH LQFHSWLRQ RI WKLV VWXG\ UHVLVWDQFH RI %UD[WRQ WR IROLDUIHHGLQJ LQVHFWV KDV EHHQ

PAGE 76

GRFXPHQWHG E\ 5RZDQ HW DO f LQ D VWXG\ WR FRPSDUH WKH UHVLVWDQFH RI UHFRPPHQGHG VR\EHDQ FXOWLYDUV LQ 0DWXULW\ *URXSV 9 9, 9,, DQG 9,,, ,Q WKDW VWXG\ %UD[WRQ ZDV FLWHG DV EHLQJ RQH RI WKH PRVW UHVLVWDQW FXOWLYDUV LQ 0DWXULW\ *URXS 9,, $V LQ WKH LQVHFWUHVLVWDQW JHUPSODVP OLQHV 3, DQG ZHUH LQFOXGHG LQ ERWK FDJHV WR SURYLGH DGGLWLRQDO FKHFNV DQG WR GRFXPHQW WKH OHYHO RI UHVLVWDQFH LQ WKH UHVLVWDQW SDUHQW ,Q ERWK FDJHV ZDV YLVLEO\ DQG VWDWLVWLFDOO\ 7DEOH f PRUH UHVLVWDQW WKDQ WKH RULJLQDO VRXUFH RI UHVLVWDQFH 3, DOVR H[KLELWHG PRUH UHVLVWDQFH WKDQ GLG LQ ERWK FDJHV DOWKRXJK LW ZDV VLJQLILFDQWO\ GLIIHUHQW DW S /6'f RQO\ LQ FDJH 7DEOH f 7KLV VXJJHVWV WKDW GXULQJ WKH ILYH F\FOHV RI FURVVLQJ DQG VHOHFWLRQ WKDW RFFXUUHG LQ WKH GHYHORSPHQW RI IURP 3, DGGLWLRQDO JHQHV ZHUH JDLQHG IURP f7UDF\0f D SDUHQW LQ WKH IRXUWK F\FOH 7UDF\0 KDV VXEVHTXHQWO\ EHHQ LGHQWLILHG DV EHLQJ PHDVXUDEO\ UHVLVWDQW WR IROLDUIHHGLQJ LQVHFWV f 6XPPDU\ DQG &RQFOXVLRQV 7KLV UHVHDUFK ZDV FRQGXFWHG WR GHWHUPLQH WKH PRGH RI LQKHULWDQFH RI UHVLVWDQFH WR VR\EHDQ ORRSHU LQ VR\EHDQ 7KH UHVXOWV IURP 0DWKHUfV VFDOLQJ WHVW f DSSOLHG WR WKH GDWD IURP HDFK FDJH VWXG\ LQGLFDWHG WKDW JHQHUDWLRQ PHDQV GHSHQG RQO\ RQ WKH DGGLWLYH JHQH HIIHFWV IRU UHVLVWDQFH WR VR\EHDQ ORRSHU 8WLOL]LQJ +D\PDQfV f PHWKRGRORJ\ LQ WHUPV RI WKH JHQHUDWLRQPHDQV DQDO\VLV WR DQDO\]H WKH GDWD IURP HDFK FDJH VHSDUDWHO\ RQO\ WKH DGGLWLYH JHQHWLF HIIHFWV ZHUH VLJQLILFDQW :KHQ WKH GDWD IURP WKH WZR FDJHV ZHUH FRPELQHG IRU DQDO\VLV WKH DG HSLVWDWLF HIIHFW ZDV VLJQLILFDQW S f +RZHYHU WKH SULPDU\ HIIHFW LV DVVXPHG WR EH DGGLWLYH EDVHG RQ WKH UHVXOWV RI WKH DQDO\VLV

PAGE 77

7DEOH $YHUDJH GHIROLDWLRQ VFRUHV LQ VR\EHDQ SRSXODWLRQV UHVLVWDQW WR VR\EHDQ ORRSHU LQ FDJH VWXG\ 3RSXODWLRQ Q 0HDQW &DJH D D 3, E &DJH D E 3, E I 0HDQV IROORZHG E\ WKH VDPH OHWWHU DUH QRW VLJQLILFDQWO\ GLIIHUHQW DW WKH S /6'f

PAGE 78

RI WKH GDWD IURP FDJH W S f FDJH W S f DQG FRPELQHG W S f DV ZDV LQGLFDWHG E\ WKH VFDOLQJ WHVW :KHQ 3RZHUfV SDUWLWLRQ DQDO\VLV f ZDV DSSOLHG WR WKH GDWD IURP HDFK FDJHV WKH ) GDWD GLG QRW ILW D WZRJHQH DGGLWLYH PRGHO IRU UHVLVWDQFH WR VR\EHDQ ORRSHU $ PRGLILFDWLRQ RI 3RHKOPDQfV f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f WR GHWHUPLQH WKH QXPEHU RI JHQHV WKDW QRW RQO\ FRQWURO UHVLVWDQFH WR VR\EHDQ ORRSHU EXW RWKHU LQVHFW VSHFLHV DV ZHOO +HULWDELOLW\ ZDV HVWLPDWHG WR EH b IRU UHVLVWDQFH WR VR\EHDQ ORRSHU %URDG VHQVH KHULWDELOLWLHV ZHUH FDOFXODWHG EHFDXVH WKH VL]H DQG YDULDQFHV RI WKH %& DQG %& SRSXODWLRQV ZHUH LQVXIILFLHQW WKXV SURKLELWLQJ FDOFXODWLRQ RI QDUURZVHQVH KHULWDELOLW\ 7KH SUHYLRXV DQDO\VHV VXJJHVWV WKDW UHVLVWDQFH WR VR\EHDQ ORRSHU LQ WKLV FURVV LV FRQWUROOHG E\ WZR JHQHV ZKLFK EHKDYH DGGLWLYHO\ WKHUHIRUH LW FDQ EH DVVXPHG WKDW WKLV KHULWDELOLW\ LV GXH RQO\ WR WKH DGGLWLYH JHQHWLF YDULDQFH 6LQFH WKH JHQHWLF YDULDWLRQ

PAGE 79

LQIOXHQFHV UHVLVWDQFH WR VR\EHDQ ORRSHU PRUH WKDQ HQYLURQPHQWDO YDULDWLRQ D EUHHGHU VKRXOG EH DEOH WR PDNH VHOHFWLRQV LQ WKH ) SRSXODWLRQ ZLWK FRQILGHQFH %\ XWLOL]LQJ D SRSXODWLRQ IURP D FURVV EHWZHHQ D UHVLVWDQW EUHHGLQJ OLQH DQG D KLJKO\ VXVFHSWLEOH EUHHGLQJ OLQH LW VKRXOG EH SRVVLEOH WR GHWHUPLQH LI WKHUH DUH RQO\ WKUHH JHQHV IRU UHVLVWDQFH WR VR\EHDQ ORRSHU LQ EUHHGLQJ OLQHV GHULYHG IURP 3, 7KH VLQJOH VHHG GHVFHQW PHWKRG RI EUHHGLQJ f FRXOG EH XVHG WR GHWHUPLQH WKH QXPEHU RI JHQHV LQYROYHG E\ DGYDQFLQJ D ODUJH SRSXODWLRQ WR WKH ) JHQHUDWLRQ DQG HYDOXDWLQJ UHSOLFDWHG KLOOV IRU UHDFWLRQ WR VR\EHDQ ORRSHU )RU H[DPSOH DVVXPLQJ WKUHH VHJUHJDWLQJ JHQH SDLUV b RI WKH ) OLQHV VKRXOG EH KRPR]\JRXV f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f PDUNHUV 5)/3 PDUNHUV KDYH EHHQ ZLGHO\ XVHG LQ RWKHU SODQW VSHFLHV WR FRQVWUXFW OLQNDJH PDSV DQG WR SODFH JHQHV WKDW FRQWURO TXDOLWDWLYH DQG TXDQWLWDWLYH WUDLWV RQWR WKHVH PDSV 5HFHQWO\ 5)/3 PDUNHUV KDYH EHHQ XVHG WR PDS SK\WRSKWKRUD UHVLVWDQFH ORFL LQ VR\EHDQ f 7KLV PHWKRGRORJ\ FRXOG EH XVHG WR PDS WKH JHQHV DQG SRVVLEO\ GHWHUPLQH DQ\ OLQNDJH JURXSV WKDW PD\ EH DVVRFLDWHG ZLWK WKH JHQHV FRQWUROOLQJ UHVLVWDQFH WR VR\EHDQ ORRSHU

PAGE 80

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f KDYH DOUHDG\ FRQGXFWHG VWXGLHV RQ WKH PHFKDQLVPV RI UHVLVWDQFH EXW IXUWKHU UHVHDUFK LV QHHGHG $V QHZ WHFKQRORJ\ EHFRPHV DYDLODEOH WKH P\VWHULHV RI WKH PHFKDQLVPV RI LQVHFW UHVLVWDQFH DQG WKH JHQHV WKDW FRQWURO WKHLU H[SUHVVLRQ DUH FORVHU WR EHLQJ VROYHG

PAGE 81

&+$37(5 6800$5< $1' &21&/86,216 6R\EHDQ LV JURZQ ZRUOGZLGH SULPDULO\ IRU LWV RLO DQG PHDO SURGXFWV DQG WR D OHVVHU GHJUHH WKH ZKROH EHDQ SURGXFW $V LQ DQ\ FURS VR\EHDQ LV H[SRVHG WR ELRWLF DQG DELRWLF \LHOGUHGXFLQJ IDFWRUV VXFK DV GLVHDVHV QHPDWRGHV LQVHFWV DQG GURXJKW WKURXJKRXW WKH JURZLQJ VHDVRQ 7R SURGXFH D SURILWDEOH FURS D JURZHU PXVW PRGLI\ WKH HIIHFWV RI WKHVH \LHOGUHGXFLQJ IDFWRUV E\ RQH RU PRUH RI WKH PHWKRGV GHVFULEHG E\ 0HWFDOI DQG /XFNPDQQ f (QYLURQPHQWDOO\ WKH EHVW DSSURDFK WR DQ\ SHVW SUREOHP ZRXOG EH WR XWLOL]H JHQHV IRU UHVLVWDQFH LQ DYDLODEOH JHUPSODVP WR GHYHORS UHVLVWDQW FXOWLYDUV f 6R\EHDQ ORRSHU LV D PDMRU SHVW WR VR\EHDQ LQ PDQ\ DUHDV RI WKH VRXWKHDVWHUQ 8QLWHG 6WDWHV 7KH VR\EHDQ ORRSHU PRUH VR WKDQ DQ\ RWKHU DUWKURSRG SHVW KDV EHFRPH LQFUHDVLQJO\ UHVLVWDQW WR DYDLODEOH LQVHFWLFLGHV f PDNLQJ LW GLIILFXOW WR FRQWURO %\ XWLOL]LQJ VR\EHDQ JHUPSODVP OLQHV WKDW SRVVHVV JHQHV IRU UHVLVWDQFH WR IROLDU IHHGLQJ LQVHFWV LW LV SRVVLEOH WR LQFUHDVH WKH KRVW SODQW UHVSRQVHV IRU FRQWURO DV KDV EHHQ GHPRQVWUDWHG E\ WKH GHYHORSPHQW RI WKH LQVHFWUHVLVWDQW FXOWLYDUV /DPDU f DQG &URFNHWW f 7KH REMHFWLYHV RI WKLV VWXG\ ZHUH Lf WR HYDOXDWH WKUHH PHWKRGV RI UDWLQJ IRU GHIROLDWLRQ RI VR\EHDQ E\ VR\EHDQ ORRSHU DQG f WR GHWHUPLQH WKH LQKHULWDQFH RI

PAGE 82

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f WKH SDUWLWLRQHG SODQW DYHUDJH YLVXDO UDWLQJ $9f DQG PHDVXUHG DYHUDJH OHDI DUHD RI WKH SDUWLWLRQHG SODQW $/f 7KH :3 DQG $9 ZHUH YLVXDO DVVHVVPHQWV RI GHIROLDWLRQ EDVHG RQ D OWR VFDOH GLYLGHG LQWR b LQFUHPHQWV ZLWK D VFRUH RI VLJQLI\LQJ GHIROLDWLRQ RI WR b DQG D VFRUH RI VLJQLI\LQJ GHIROLDWLRQ RI WR b 7KH :3 PHWKRG ZDV EDVHG RQ RQH REVHUYDWLRQ SHU SODQW ZKHUHDV WKH $9 PHWKRG ZDV EDVHG RQ WKH PHDQ RI WKUHH REVHUYDWLRQV SHU SODQWV 7KH DEVROXWH PHWKRG $/f XWLOL]HG D /,&25 /, OHDI DUHD PHWHU WR PHDVXUH WKH DUHD RI WKUHH WULIROLRODWHV RQH HDFK IURP WKH WRS PLGGOH DQG ERWWRP RI WKH SODQW ,Q QR DEVROXWH PHWKRG FRPSDULVRQV ZHUH SRVVLEOH KRZHYHU WKH :3 DQG $9 ZHUH KLJKO\ FRUUHODWHG U S f $GMXVWPHQWV LQ WKH PHWKRG RI FDOFXODWLQJ SHUFHQW GHIROLDWLRQ E\ $/ IRU DOORZHG IRU GLUHFW FRPSDULVRQV ZLWK WKH :3 DQG $9 PHWKRGV $OO FRUUHODWLRQV :3 YV $9 :3 YV $/ DQG $9 YV $/f ZHUH VLJQLILFDQW DW WKH S OHYHO ZKHWKHU WKH FDJHV ZHUH H[DPLQHG VHSDUDWHO\ RU SRROHG 7KH UHVXOWV RI WKHVH FRUUHODWLRQV LQGLFDWH WKDW WKH HYDOXDWRU FRXOG REWDLQ VLPLODU HVWLPDWHV RI GHIROLDWLRQ UHJDUGOHVV RI ZKLFK RI WKH WKUHH PHWKRGV LV XWLOL]HG 7KHVH ILQGLQJV DUH

PAGE 83

VLPLODU WR WKH UHVXOWV RI DQRWKHU UHODWLYH YHUVXV DEVROXWH HVWLPDWLRQ RI GHIROLDWLRQ VWXG\ FRQGXFWHG LQ E\ *5 %RZHUV *5 %RZHUV SHUVRQDO FRPPXQLFDWLRQf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fV HVWLPDWHV RI GHIROLDWLRQ ZLWK WKH :3 PHWKRG ZHUH FORVHU WR WKH DFWXDO WKDQ ZHUH WKH HVWLPDWHV RI WKH $9 PHWKRG 7KHVH ILQGLQJV ZHUH YDOLGDWHG ZLWK 0F1HPDUfV WHVW IRU FRQFRUGDQFH 7KH UHODWLYH YDULDWLRQ 59f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

PAGE 84

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fV VFDOLQJ WHVW f ZDV DSSOLHG WR WKH GHIROLDWLRQ GDWD IURP HDFK FDJH VWXG\ DQG LW ZDV GHWHUPLQHG WKDW JHQHUDWLRQ PHDQV GHSHQG RQO\ RQ WKH DGGLWLYH JHQH HIIHFWV DQG QR HSLVWDVLV ZDV SUHVHQW 8WLOL]LQJ +D\PDQfV f PHWKRGRORJ\ LQ WHUPV RI WKH

PAGE 85

JHQHUDWLRQPHDQV DQDO\VLV WR DQDO\]H WKH GDWD IURP HDFK FDJH VHSDUDWHO\ RQO\ WKH DGGLWLYH JHQHWLF HIIHFWV ZHUH VLJQLILFDQW :KHQ WKH GDWD IURP WKH WZR FDJHV ZHUH FRPELQHG IRU DQDO\VLV WKH DG HSLVWDWLF HIIHFW ZDV VLJQLILFDQW ^W S f +RZHYHU WKH SULPDU\ HIIHFW LV DVVXPHG WR EH DGGLWLYH EDVHG RQ WKH UHVXOWV RI WKH DQDO\VLV RI WKH GDWD IURP FDJH W S f FDJH W S f DQG FRPELQHG W S f DV ZDV LQGLFDWHG E\ WKH VFDOLQJ WHVW $SSO\LQJ 3RZHUfV SDUWLWLRQLQJ DQDO\VLV f WR WKH GDWD IURP HDFK FDJH LW ZDV GHWHUPLQHG WKDW LQ WKLV FURVV WKH ) GDWD GLG QRW ILW WKH WZRJHQH DGGLWLYH PRGHO IRU UHVLVWDQFH WR VR\EHDQ ORRSHU 7KHUH LV JRRG UHDVRQ WR EHOLHYH WKDW WKH GHIROLDWLRQ VFRUHV DVVLJQHG LQ WKLV H[SHULPHQW VKRXOG QRW EH DVVXPHG WR FRUUHVSRQG WR JHQRW\SH HIIHFWV WKHUHIRUH WKH 3RZHUfV SDUWLWLRQLQJ DQDO\VLV KDV QR XWLOLW\ LQ WKHVH DQDO\VHV %DVHG RQ D PRGLILFDWLRQ RI WKH IRUPXOD SUHVHQWHG E\ 3RHKOPDQ f LW ZDV GHWHUPLQHG WKDW WZR JHQHV FRQWULEXWH WR WKH H[SUHVVLRQ RI UHVLVWDQFH WR VR\EHDQ ORRSHU LQ WKLV FURVV +RZHYHU WKLV IRUPXOD RQO\ HVWLPDWHV WKH QXPEHU RI JHQHV IRU UHVLVWDQFH WR VR\EHDQ ORRSHU E\ ZKLFK WKH WZR SDUHQWV GLIIHU ,Q DFWXDOLW\ WKH QXPEHU RI JHQHV IRU UHVLVWDQFH WR VR\EHDQ ORRSHU LV SUREDEO\ JUHDWHU WKDQ WZR EXW LW ZRXOG WDNH IXUWKHU UHVHDUFK WR GHWHUPLQH WKLV %URDG VHQVH KHULWDELOLW\ EDVHG RQ WKH GHIROLDWLRQ GDWD ZDV HVWLPDWHG WR EH b IRU UHVLVWDQFH WR VR\EHDQ ORRSHU 6LQFH WKH SUHYLRXV DQDO\VHV VXJJHVW WKDW UHVLVWDQFH WR WKH VR\EHDQ ORRSHU LV FRQWUROOHG E\ WZR JHQHV ZKLFK EHKDYH DGGLWLYHO\ LW FDQ EH DVVXPHG WKDW WKLV KHULWDELOLW\ LV GXH RQO\ WR WKH DGGLWLYH JHQHWLF YDULDWLRQ $ KHULWDELOLW\ RI WKLV

PAGE 86

PDJQLWXGH VXJJHVWV WKDW D EUHHGHU VKRXOG EH DEOH WR PDNH SURJUHVV E\ VHOHFWLQJ IRU UHVLVWDQFH WR VR\EHDQ ORRSHU LQ HDUO\ JHQHUDWLRQV 7KLV UHVHDUFK FRXOG EH FRQWLQXHG ZLWK EUHHGLQJ OLQHV GHULYHG IURP 3, RU E\ LQFRUSRUDWLQJ DGGLWLRQDO VRXUFHV RI UHVLVWDQFH VXFK DV 3, 3, 3, RU 3, LQWR DJURQRPLFDOO\ DGDSWHG EUHHGLQJ OLQHV 7KHVH HOLWH JHQRW\SHV FRXOG WKHQ EH FURVVSROOLQDWHG ZLWK D KLJKO\ VXVFHSWLEOH FXOWLYDU VXFK DV &HQWHQQLDO WR GHYHORS ODUJH ) SRSXODWLRQV IRU IXWXUH HYDOXDWLRQV ,Q RUGHU WR GHWHUPLQH WKH QXPEHU RI JHQHV FRQWUROOLQJ UHVLVWDQFH WR WKH VR\EHDQ ORRSHU DQGRU RWKHU VSHFLHV WKHVH ) SRSXODWLRQV FRXOG EH VFUHHQHG E\ FRQYHQWLRQDO PHWKRGV RU VRPH RI WKH QHZHU PHWKRGV VXFK DV 5)/3 DQDO\VLV 6LQFH UHVLVWDQFH WR VR\EHDQ ORRSHU LQ WKLV VWXG\ ZDV GHWHUPLQHG WR EH DGGLWLYH LW VKRXOG EH SRVVLEOH WR LQFUHDVH UHVLVWDQFH LQ IXWXUH FXOWLYDUV E\ S\UDPLGLQJ DGGLWLRQDO JHQHV IURP RWKHU VRXUFHV RI UHVLVWDQFH 7KLV FDQ EH GRQH E\ FRQYHQWLRQDO SODQW EUHHGLQJ WHFKQLTXHV E\ XWLOL]LQJ VR\EHDQ JHUPSODVP OLQHV WKDW VKRZ UHVLVWDQFH WR OHDIIHHGLQJ LQVHFWV DQG LQFRUSRUDWLQJ WKHVH JHQHV LQWR DGDSWHG FXOWLYDUV RU DGYDQFHG EUHHGLQJ OLQHV 7KH VDPH JRDOV FRXOG DOVR EH DFFRPSOLVKHG E\ DQ\ RI WKH QHZ PHWKRGV RI ELRWHFKQRORJ\ WKDW DUH FXUUHQWO\ DYDLODEOH $GGLWLRQDO UHVHDUFK ZLOO QHHG WR EH FRQGXFWHG RQ VRXUFHV RI UHVLVWDQFH RWKHU WKDQ 3, GHULYHG OLQHV WR GHWHUPLQH KRZ PDQ\ RWKHU JHQHV IRU UHVLVWDQFH DUH DYDLODEOH WR EUHHGHUV IRU XWLOL]DWLRQ LQ WKHLU EUHHGLQJ SURJUDPV ,Q RUGHU WR IXOO\ H[SORLW WKH EHQHILWV RI LQVHFW UHVLVWDQFH LQ D EUHHGLQJ SURJUDP LW ZLOO EHFRPH QHFHVVDU\ WR XQGHUVWDQG WKH PHFKDQLVPV RI UHVLVWDQFH DV ZHOO DV WKH XQGHUO\LQJ JHQHV WKDW FRGH IRU WKH

PAGE 87

PHFKDQLVP $V QHZ WHFKQRORJ\ EHFRPHV DYDLODEOH WKH P\VWHULHV RI WKH PHFKDQLVPV RI LQVHFW UHVLVWDQFH ZLOO EH VROYHG ,GHDOO\ WKH IXWXUH SURPLVHV WKH GHYHORSPHQW RI D UHVLVWDQW FXOWLYDU WKDW KDV PDQ\ JHQHV IRU UHVLVWDQFH WKDW FRGH IRU DOO WKUHH PHFKDQLVPV RI UHVLVWDQFH

PAGE 88

$33(1',; $ /,67 2) '()2/,$7,21 6&25(6 *(1(5$7(' %< :+2/( 3/$17 9,68$/ 3$57,7,21(' 3/$17 9,68$/ $1' /($) $5($ 2) 7+( 3$57,7,21(' 3/$17 5$7,1* 0(7+2'6 ,1

PAGE 89

/LVW RI GHIROLDWLRQ VFRUHV JHQHUDWHG E\ ZKROH SODQW YLVXDOI SDUWLWLRQHG SODQW YLVXDOI DQG OHDI DUHD RI WKH SDUWLWLRQHG SODQWV UDWLQJ PHWKRGV :KROH $YJ $YJ 2EVHUYDWLRQ 3ODQW 3ODQW 9LVXDO /HDI $UHD %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%(

PAGE 90

3ODQW :KROH 3ODQW $YJ 9LVXDO $YJ /HDI $UHD %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%(

PAGE 91

3ODQW :KROH 3ODQW $YJ 9LVXDO $YJ /HDI $UHD %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%(

PAGE 92

2EVHUYDWLRQ 3ODQW :KROH 3ODQW $YJ 9LVXDO $YJ /HDI $UHD %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%(

PAGE 93

2EVHUYDWLRQ 3ODQW :KROH 3ODQW $YJ 9LVXDO $YJ /HDI $UHD %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%(

PAGE 94

2EVHUYDWLRQ 3ODQW :KROH 3ODQW $YJ 9LVXDO $YJ /HDI $UHD %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( %%( 0HDQ Q f 9DULDQFH 6WG'HY W %DVHG RQ D WR VFDOH GLYLGHG LQWR b LQFUHPHQWV ZLWK D UHIOHFWLQJ b GHIROLDWLRQ /,&25 /, SRUWDEOH OHDI DUHD PHWHU PHDVXUHPHQWV LQ FP RI OHDI WLVVXH

PAGE 95

$33(1',; % /,67 2) '()2/,$7,21 6&25(6 *(1(5$7(' %< :+2/( 3/$17 9,68$/ 3$57,7,21(' 3/$17 9,68$/ $1' /($) $5($ 2) 7+( 3$57,7,21(' 3/$17 5$7,1* 0(7+2'6 ,1

PAGE 96

/LVW RI GHIROLDWLRQ VFRUHV JHQHUDWHG E\ ZKROH SODQW YLVXDO SDUWLWLRQHG SODQW YLVXDO DQG OHDI DUHD RI WKH SDUWLWLRQHG SODQW UDWLQJ PHWKRGV 2EVHUYDWLRQ &DJH 3ODQW :KROH SODQW $YJ YLVXDO $YJ /,&25 i

PAGE 97

2EVHUYDWLRQ &DJH 3ODQW :KROH SODQW $YJ YLVXDO $YJ /,&25

PAGE 98

2EVHUYDWLRQ &DJH 3ODQW :KROH SODQW $YJ YLVXDO $YJ /,&25 0HDQ 1 f 9DULDQFH 6WG'HY W %DVHG RQ D WR VFDOH GLYLGHG LQWR b LQFUHPHQWV ZLWK D UHIOHFWLQJ b GHIROLDWLRQ %DVHG RQ /,&25 /, SRUWDEOH OHDI DUHD PHWHU PHDVXUHPHQWV WKDW KDYH EHHQ FRQYHUWHG WR b GHIROLDWLRQ DQG FDWHJRUL]HG LQWR WKH OWR VFDOH i 0LVVLQJ GDWD SRLQWV

PAGE 99

$33(1',; & *(1(5$7,21 0($16 $1$/<6,6 6$6 352*5$0 237,216 /6 36 r7KLV LV D *HQHUDWLRQ 0HDQV $QDO\VLV EDVHG RQ +D\PDQfV PHWKRGRORJ\ +HUHGLW\ r '$7$ ,1),/( f$U&DJHOGDWf r 7KH ,1387 VWDWHPHQW ZLOO YDU\ DFFRUGLQJ WR WKH GDWD VHW \RX QHHG JHQHUDWLRQ *(1f DQG D GHSHQGHQW YDULDEOH r ,1387 URZ SODQW JHQ IF UDWLQJ UDWLQJ UDWLQJ 0($15$7( UDWLQJ UDWLQJUDWLQJf 352& 6257 %< *(1 352& 0($16 %< *(1 9$5 0($15$7( 287387 287 1( 0($1 < 9$5 9 67'(55 6 '$7$ 1(: 6(7 1( 56 6 ,) *(1 f'f 25 *(1 f%'f 7+(1 '(/(7( 352& 35,17 r
PAGE 100

r 7KLV PRGHO WHVWV WKH VLJQLILFDQFH RI WKH DGGLWLYH ;,f DQG WKH GRPLQDQW ;f HIIHFWV r 7,7/( f352& 5(* :(,*+7('f 352& 5(* 02'(/ < ;, ; :(,*+7 56 r 7KH PRGHO PXVW EH ZHLJKWHG WR DFFRXQW IRU WKH XQHTXDO SRSXODWLRQ VL]HV DPRQJ WKH JHQHUDWLRQV 6HH 5RZH DQG $OH[DQGHU &URS 6FL IRU IXUWKHU GHWDLO 7KH QH[W PRGHO WHVWV IRU WKH HSLVWDWLF HIIHFWV DGGLWLYHDGGLWLYH ;f DGGLWLYHGRPLQDQW ;f DQG GRPLQDQWGRPLQDQW ;f DV ZHOO DV WKH DGGLWLYH ;,f DQG GRPLQDQW ;f HIIHFWV r 352 5(* 02'(/ < ;, ; ; ; ; :(,*+7 56 581 r 7KLV QH[W VHW RI PRGHOV DUH WHVWHG ZLWK 352& */0 LQVWHDG RI 352& 5(* r 7,7/( f352& */0 :(,*+7('f 352& */0 02'(/ < ;, ; :(,*+7 56 352& */0 02'(/ < ;, ; ; ; ; :(,*+7 56 581 r 7KH QH[W VHULHV RI PRGHOV DUH WKH VDPH DV WKH SUHYLRXV PRGHOV H[FHSW WKDW WKH\ DVVXPH HTXDO SRSXODWLRQ VL]H DPRQJ WKH JHQHUDWLRQV DQG DUH WKHUHIRUH QRW ZHLJKWHG r 7,7/( f352& 5(* 12:(,*+7f 352& 5(* 02'(/ < ;, ; 352& 5(* 02'(/ < ;, ; ; ; ; 581 7,7/( f352& */0 12:(,*+7f 02'(/ < ;, ; 352& */0 02'(/ < ;, ; ; ; ; 581 r 'HSHQGLQJ RQ ZKHWKHU RU QRW WKH SRSXODWLRQ VL]HV DPRQJ WKH JHQHUDWLRQV DUH HTXDO UHSRUW WKH PRGHO WKDW KDV WKH PRVW VLJQLILFDQW HIIHFWV IURP HLWKHU 352& */0 RU 352& 5(* r

PAGE 101

$33(1',; *(1(5$7,21 0($16 $1$/<6,6 6$6 352*5$0 r 7KLV LV D *HQHUDWLRQ 0HDQV $QDO\VLV WKDW LV EDVHG RQ +D\PDQfV PHWKRGRORJ\ +HUHGLW\ 7KLV DQDO\VLV WHVWV IRU EORFN HIIHFWV DV ZHOO DV HVWLPDWLQJ WKH DGGLWLYH GRPLQDQW DQG HSLVWDWLF HIIHFWV r '$7$ $ ,1),/( f$&$*('$7f&$*( r 7KH ,1387 VWDWHPHQW ZLOO YDU\ DFFRUGLQJ WR WKH GDWD VHW \RX QHHG JHQHUDWLRQ *(1f DQG D GHSHQGHQW YDULDEOH r ,1387 52: 3/$17 *(1 )& 5$7,1* 5$7,1* 5$7,1* 0($15$7( 5$7,1* 5$7,1*5$7,1*f '$7$ % ,1),/( f$&$*('$7f&$*( ,1387 52: 3/$17 *(1 )& 5$7,1* 5$7,1* 5$7,1* 0($15$7( 5$7,1* 5$7,1* 5$7,1*f '$7$ %27+6(7 $ %581 352& 6257 %< &$*( *(1 352& 0($16 %< &$*( *(1 9$5 0($15$7( 287387 287 1( 0($1 < 9$5 9 67'(55 6 '$7$ 1(: 6(7 1( 56 O6,)*(1 f'f25 *(1 f%'f 7+(1 '(/(7( 352& 35,17 r
PAGE 102

3, 3 %& %& %6 %6 ), ) ) 3, 3 352& 6257 '$7$ &2()&176%< &$*( *(1 '$7$ ),1$/ 0(5*( 1(: &2()&176%< &$*( *(1 352& 35,17 r 7KLV PRGHO WHVWV WKH VLJQLILFDQFH RI WKH DGGLWLYH ;,f DQG GRPLQDQW ;f HIIHFWV r 352& */0 '$7$ ),1$/%< &$*( 02'(/ < ;, ; 66, 66 :(,*+7 56 r 7KH PRGHO PXVW EH ZHLJKWHG WR DFFRXQW IRU WKH XQHTXDO SRSXODWLRQ VL]HV DPRQJ WKH JHQHUDWLRQV 6HH 5RZH DQG $OH[DQGHU &URS 6FL IRU IXUWKHU GHWDLO r 352& */0 '$7$ ),1$/&/$66 &$*( 02'(/ < &$*( ;, ; &$*(r; &$*(r;66 66 62/87,21 :(,*+7 56 352& */0 '$7$ ),1$/&/$66 &$*( 02'(/ < &$*( ;, ; 66, 66 62/87,21 :(,*+7 56 r 7KH QH[W PRGHO WHVWV IRU WKH HSLVWDWLF HIIHFWV DGGLWLYHDGGLWLYH ;f DGGLWLYHGRPLQDQW ;f DQG GRPLQDQWGRPLQDQW ;f DV ZHOO DV WKH DGGLWLYH ;,f DQG GRPLQDQW ;f HIIHFWV r 352& */0%< &$*( 02'(/ < ;, ; ; ; ; :(,*+7 56 581 352& */0&/$66 &$*( 02'(/ < &$*( ;, ; ; ; ; &$*(r; &$*(r; &$*(r; &$*(r; &$*(r;66 66 62/87,21 :(,*+7 56 581 352& */0&/$66 &$*( 02'(/ \ &$*( ;, ; ; ; ; 66, 66 62/87,21 :(,*+7 56 581

PAGE 103

$33(1',; ( 7(67 )25 1250$/,7< [ >f f f f f f f f@ V >Of f f f f f f Of@Of V fr 1H[W WKH YDOXH [ ZDV FDOFXODWHG IRU HDFK FODVV DV IROORZV &ODVV [ &ODVV [ &ODVV [ &ODVV [ &ODVV [ &ODVV [ &ODVV [ &ODVV [ f f f f f f f f 7KH FRUUHVSRQGLQJ = YDOXH IRU HDFK FODVV ZDV REWDLQHG E\ ILQGLQJ [ LQ D WDEOH RI SUREDELOLWLHV RI WKH QRUPDO GLVWULEXWLRQ f 7KH = YDOXH REWDLQHG IRU HDFK FODVV ZDV WKHQ VXEWUDFWHG IURP RQH DQG PXOWLSOLHG E\ b WR REWDLQ WKH WKHRUHWLFDO b RI WKH SRSXODWLRQ WKDW IHOO LQWR HDFK FODVV 7KH WKHRUHWLFDO QXPEHU RI HDFK FODVV ZDV REWDLQHG E\ PXOWLSO\LQJ WKH WRWDO Q RI WKH SRSXODWLRQ E\ WKH WKHRUHWLFDO b RI HDFK FODVV 7KLV ZDV GRQH IRU HDFK FODVV DV VKRZQ EHORZ

PAGE 104

; = = 7KHRUHWLFDO b &ODVV 7KHRUHWLFDO 'LVWULEXWLRQ 7KH REVHUYHG YHUVXV WKHRUHWLFDO GLVWULEXWLRQ RI WKH ) SRSXODWLRQ ZDV WKHQ WHVWHG ZLWK D ; WHVW

PAGE 105

5()(5(1&(6 $OODUG 5: 3ULQFLSOHV RI SODQW EUHHGLQJ -RKQ :LOH\ DQG 6RQV ,QF 1HZ
PAGE 106

'HUHQ &: DQG .+ 4XHVHQEHUU\ ,QKHULWDQFH RI SKRWRSHULRGLQGXFHG IORZHULQJ LQ WKUHH SKRWRSHULRGLF OLQHV RI $HVFK\QRPHQH DPHULFDQD / 7KHRU $SSO *HQHW 'LHUV %: / 0DQVXU ,PVDQGH DQG 5& 6KRHPDNHU 0DSSLQJ SK\WRSKWKRUD UHVLVWDQFH ORFL LQ VR\EHDQ ZLWK UHVWULFWLRQ IUDJPHQW OHQJWK SRO\PRUSKLVP PDUNHUV &URS 6FL (QQLV :% -U 0RGHP PHWKRGV IRU FRQWUROOLQJ SHVWV S ,Q /' +LOO HGf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f LQ 0LVVLVVLSSL (FRQ (QWRPRO )XQGHUEXUN -( $5 6RIIHV 5' %DUQHWW '& +HU]RJ DQG +LQVRQ 3ORW VL]H DQG VKDSH LQ UHODWLRQ WR VR\EHDQ UHVLVWDQFH IRU YHOYHWEHDQ FDWHUSLOODU /HSLGRSWHUD1RFWXLGDHf (FRQ (QWRPRO )XQGHUEXUN -( '/ :ULJKW DQG ,' 7HDUH 3UHSODQW WLOODJH HIIHFWV RQ SRSXODWLRQ G\QDPLFV RI VR\EHDQ LQVHFW SHVWV &URS 6FL *DPEOH (( *HQH HIIHFWV LQ FRP =HD PD\V /f ,VHSDUDWLRQ DQG UHODWLYH LPSRUWDQFH RI JHQH HIIHFWV IRU \LHOG &DQDGLDQ 3ODQW 6FL *DU\ '/ /DPEHUW DQG -' 2X]WV (YDOXDWLRQ RI VR\EHDQ SODQW LQWURGXFWLRQV IRU UHVLVWDQFH WR IROLDU IHHGLQJ LQVHFWV 06 $FDGHP\ 6FL +DOODXHU $5 DQG -% 0LUDQGD 4XDQWLWDWLYH JHQHWLFV LQ PDL]H EUHHGLQJ ,RZD 6WDWH 8QLY 3UHVV $PHV ,$

PAGE 107

+DPHU -/ 6R\EHDQ ORRSHU ELRORJ\ DQG DSSURDFKHV IRU LPSURYHG PDQDJHPHQW 0LVVLVVLSSL &RRSHUDWLYH ([WHQVLRQ 6HUYLFH ,QIRUPDWLRQ 6KHHW +DUW 69 -: %XUWRQ DQG :9 &DPSEHOO &RPSDULVRQ RI WKUHH WHFKQLTXHV WR HYDOXDWH DGYDQFHG EUHHGLQJ OLQHV RI VR\EHDQ IRU OHDIIHHGLQJ UHVLVWDQFH WR FRP HDUZRUP /HSLGRSWHUD1RFWXLGDHf (FRQ (QWRPRO +DUWOH\ ** 0XOWLFHOOXODU UHDULQJ PHWKRGV IRU WKH EHHW DUP\ZRUP VR\EHDQ ORRSHU DQG YHOYHWEHDQ FDWHUSLOODU /HSLGRSWHUD1RFWXLGDHf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f -HQQLQJV &: :$ 5XVVHOO :' *XWKULH DQG 5/ *ULQGHODQG *HQHWLFV RI UHVLVWDQFH LQ PDL]H WR VHFRQGEURRG (XURSHDQ FRP ERUHU ,RZD 6WDWH 5HVHDUFK -RVKL -0 DQG 0 1REDNKW (YDOXDWLRQ RI FRPPHUFLDO VR\EHDQ FXOWLYDUV DQG DGYDQFH EUHHGLQJ OLQHV IRU QRQSUHIHUHQFH WR +HOLRWKLV ]HD 6R\EHDQ *HQHWLFV 1HZVOHWWHU .LOHQ 7& -+ +DWFKHWW DQG (( +DUWZLJ (YDOXDWLRQ RI HDUO\ JHQHUDWLRQ VR\EHDQV IRU UHVLVWDQFH WR VR\EHDQ ORRSHU &URS 6FL

PAGE 108

.LOHQ 7& DQG / /DPEHUW (YLGHQFH IRU GLIIHUHQW JHQHV FRQWUROOLQJ LQVHFW UHVLVWDQFH LQ WKUHH VR\EHDQ JHQRW\SHV &URS 6FL .RJDQ 0 3ODQW UHVLVWDQFH LQ SHVW PDQDJHPHQW S ,Q 5/ 0HWFDOI DQG :+ /XFNPDQQ HGf ,QWURGXFWLRQ WR LQVHFW SHVW PDQDJHPHQW QG HG -RKQ :LOH\ DQG 6RQV 1HZ
PAGE 109

/HRQDUG :+ +2 0DQQ DQG / 3RZHUV 3DUWLWLRQLQJ PHWKRG RI JHQHWLF DQDO\VLV DSSOLHG WR SODQWKHLJKW LQKHULWDQFH LQ EDUOH\ &ROR $JULH ([S 6WQ 7HFK %XOO /XFNPDQQ :+ DQG 5/ 0HWFDOI 7KH SHVW PDQDJHPHQW FRQFHSW S ,Q 5/ 0HWFDOI DQG :+ /XFNPDQQ HGf ,QWURGXFWLRQ WR LQVHFW SHVW PDQDJHPHQW QG HG -RKQ :LOH\ DQG 6RQV 1HZ
PAGE 110

5RVH 5/ %5 /HRQDUG 7& 6SDUNV DQG -% *UDYHV (QKDQFHG PHWDEROLVP DQG NQRFNGRZQ UHVLVWDQFH LQ D ILHOG YHUVXV D ODERUDWRU\ VWUDLQ RI WKH VR\EHDQ ORRSHU /HSLGRSWHUD 1RFWXLGDHf (FRQ (QWRPRO 5RVH 5/ 7& 6SDUNV DQG &0 6PLWK 7KH LQIOXHQFH RI UHVLVWDQW VR\EHDQ 3, f IROLDJH DQG FRXPHVWURO RQ WKH PHWDEROLVP RI [HQRELRWLFV E\ WKH VR\EHDQ ORRSHU 3VHXGRSOXVLD LQFOXGHQV :DONHUf 3HVW %LRFKHP DQG 3K\VLRO 5RZDQ *% +5 %RHUPD -1 $OO DQG 7RGG 6R\EHDQ FXOWLYDU UHVLVWDQFH WR GHIROLDWLQJ LQVHFWV &URS 6FL 5RZH .( DQG :/ $OH[DQGHU &RPSXWDWLRQV IRU HVWLPDWLQJ WKH JHQHWLF SDUDPHWHUV LQ MRLQWVFDOLQJ WHVWV &URS 6FL 5XHVLQN :* ,QWURGXFWLRQ WR VDPSOLQJ WKHRU\ S ,Q 0 .RJDQ DQG '&+HU]RJ HGVf 6DPSOLQJ PHWKRGV LQ VR\EHDQ PHWKRGRORJ\ 6SULQJHU 1HZ
PAGE 111

6PLWK &0 3ODQW UHVLVWDQFH WR LQVHFWV D IXQGDPHQWDO DSSURDFK -RKQ :LOH\ DQG 6RQV 1HZ
PAGE 112

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

PAGE 113

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

PAGE 114

7KLV GLVVHUWDWLRQ ZDV VXEPLWWHG WR WKH *UDGXDWH )DFXOW\ RI WKH &ROOHJH RI $JULFXOWXUH DQG WR WKH *UDGXDWH 6FKRRO DQG ZDV DFFHSWHG DV SDUWLDO IXOILOOPHQW RI WKH UHTXLUHPHQWV IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ $SULO 'HDQ &ROOHJH RI $JULFXOWXUH 'HDQ *UDGXDWH 6FKRRO

PAGE 115

/ ILOR L A mDr 81,9(56,7< 2) )/25,'$


INHERITANCE OF RESISTANCE TO THE SOYBEAN LOOPER
IN SOYBEAN
By
MICHAEL MONTGOMERY KENTY
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1994

I would like to dedicate this work to the memory of the late Dr. Johnny D. Ouzts.
Thank you for your friendship and guidance; without it I would not have made it this far.

ACKNOWLEDGEMENTS
My deepest gratitude is extended to Dr. Kuell Hinson for serving as my mentor
throughout my program of study. I would also like to thank the remaining members of
my graduate committee, Drs. Joe Funderburk, Ken Quesenberry, John Strayer, and
David Wofford, for their insight and guidance. This study could not have been
completed without the support and aid of the USDA-ARS Soybean Production Research
Unit at Stoneville, Mississippi. I want to thank each and every member for their
friendship, patience, and support during this educational process. Also, I would like to
thank Dr. Clarence Watson and Debbie Boykin for their statistical advice. A special
thanks goes to Drs. Edgar E. Hartwig and Thomas C. Kilen for taking a chance on a
young kid years ago, and introducing me to the wonderful field of plant breeding.
in

TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS iii
LIST OF TABLES vi
LIST OF FIGURES viii
ABSTRACT ix
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 LITERATURE REVIEW 5
CHAPTER 3 EVALUATION OF RATING METHODS 13
Introduction 13
Literature Review 13
Materials and Methods 15
Results and Discussion 19
Summary and Conclusions 29
CHAPTER 4 INHERITANCE OF RESISTANCE TO SOYBEAN LOOPER IN
SOYBEAN 31
Introduction 31
Materials and Methods 32
1988 32
1989 33
1990 34
1991 39
Results and Discussion 49
1990 49
1991 51
Summary and Conclusions 66
CHAPTER 5 SUMMARY AND CONCLUSIONS 71
Evaluation of Rating Methods 72
Inheritance Study 74
IV

APPENDIX A LIST OF DEFOLIATION SCORES GENERATED BY
WHOLE PLANT VISUAL, PARTITIONED PLANT VISUAL, AND
LEAF AREA OF THE PARTITIONED PLANT RATING METHODS
IN 1990 78
APPENDIX B LIST OF DEFOLIATION SCORES GENERATED BY
WHOLE PLANT VISUAL, PARTITIONED PLANT VISUAL, AND
LEAF AREA OF THE PARTITIONED PLANT RATING METHODS
IN 1991 85
APPENDIX C GENERATION MEANS ANALYSIS SAS PROGRAM 89
APPENDIX D GENERATION MEANS ANALYSIS SAS PROGRAM 91
APPENDIX E TEST FOR NORMALITY 93
REFERENCES 95
BIOGRAPHICAL SKETCH 102
v

LIST OF TABLES
page
Table 1. McNemar test for significant differences between the WP (whole plant
visual) and AV (partitioned plant visual) methods in 1990; and between
the WP and AV, WP and AL (leaf area of the partitioned plant), and AV
and AL methods in 1991 28
Table 2. Populations used for generation means analysis of the inheritance of
resistance to the soybean looper 46
Table 3. Ratings of leaf feeding by soybean looper on soybean parents, their F,
and F2 progeny, and the germplasm lines PI 229358 and D75-10169 in the
field cage-1990 50
Table 4. Goodness of fit test for normality! of soybean looper defoliation for
populations of D86-3429, Braxton, and D86-3429 x Braxton F2
populations in the 1990 cage study 52
Table 5. Ratings of leaf feeding by soybean looper on soybean parents, their F1;
F2, F3, BC,, BSl5 BC2, and BS2 progeny, and the germplasm lines PI
229358 and D75-10169 in the field cages-1991 54
Table 6. Estimates of variance components obtained from an analysis of the
homogeneous populations, D86-3429, Braxton, D75-10169, PI 229358,
for a location effect due to feeding+ by soybean looper in field cages in
1991 55
Table 7. Means, standard errors, and variances of soybean looper defoliation
ratings for parental lines and seven populations of soybean arising from
the cross D86-3429 x Braxton grown in field cages at Stoneville, MS in
1991 57
Table 8. Mather’s scaling test applied to the cross D86-3429 x Braxton to test
adequacy of additive-dominance model for resistance to soybean looper. . 58
vi

Table 9. Estimates of the additive, dominant, and epistatic effects in the
generation means for defoliation by soybean looper in the nine populations
of the D86-3429 x Braxton material grown in the field cages at
Stoneville, MS in 1991 60
Table 10. Estimates of the additive, dominant, and epistatic effects in the
generation means for defoliation by soybean looper in the nine populations
of the D86-3429 x Braxton material grown in the field cages at
Stoneville, MS in 1991 61
Table 11. Goodness of fit test for normality of soybean looper defoliation data
from 1991 cage study 63
Table 12. Average defoliation scores in soybean populations resistant to soybean
looper in 1991 cage study 67
Vll

LIST OF FIGURES
E2ge
Figure 1. Examples of photocopied leaflets used in determining the % defoliation
in the AV (leaf area of partitioned plant) method 17
Figure 2. 1990 frequency distribution of defoliation estimates by the whole plant
and partitioned plant average visual rating methods 21
Figure 3. 1991 frequency distribution of defoliation estimates by the whole plant,
partitioned plant average visual, and measured average leaf area of the
partitioned plant rating methods 22
Figure 4. Schematic drawing of field cages used in the 1990 and 1991
experiment 35
Figure 5. Mating scheme to derive populations required to estimate the genetic
effects of resistance to the soybean looper in soybean 45
Vlll

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of
Doctor of Philosophy
INHERITANCE OF RESISTANCE TO THE SOYBEAN LOOPER
IN SOYBEAN
By
Michael Montgomery Kenty
April, 1994
Chairman: Kuell Hinson
Major Department: Agronomy
Phytophagous insects cause millions of dollars of damage to soybean [Glycine max
(L.) Merr.] throughout the southern United States. This study was conducted to: (/)
evaluate three methods for rating defoliation of soybean by insects and (ii) determine the
inheritance of resistance to the soybean looper [Pseudoplusia includens Walker] in
soybean. The insect-resistant breeding line D86-3429 was crossed with the cultivar
Braxton in 1988. Additional crosses were made in 1989 and 1990 to provide the
populations necessary to conduct rating methodology and genetic studies in field cages
in 1990 and 1991. The three rating methods evaluated were whole plant visual (WP),
partitioned plant average visual (AV), and measured average leaf area of partitioned plant
(AL). In 1990 the WP and AL were highly correlated (r=0.65, p < 0.001) and in 1991
all correlations (WP vs. AV, WP vs. AL, and AV vs. AL) were significant at p <
IX

0.001. Based on the relative variation of each method, the precision of the three
methods was essentially equal and very good, well below the desirable level of 10.
These results indicated that relative estimates are suitable for genetic studies. In the
inheritance study defoliation was estimated by the WP method in 1990 and the AV
method in 1991. The data from the preliminary study in 1990 showed a trend towards
quantitative inheritance, therefore the 1991 data were analyzed quantitatively. Mather’s
scaling test was applied to the data generated from each cage and the results indicated
that generation means depend only on additive gene effects. Utilizing Hayman’s
methodology in terms of the generation means analysis, an epistatic effect was suggested,
but the primary effect is assumed to be additive, as was indicated in the scaling test.
Estimates of gene numbers indicate that the two parents differed by two genes for
resistance to soybean looper. Heritability for resistance was estimated to be 63%,
therefore a breeder should be able to make progress by selecting in the F2 or F3
populations. The potential exists for increasing resistance to the soybean looper, and
possibly other phytophagous insects, by pyramiding genes from new sources of
resistance.
x

CHAPTER 1
INTRODUCTION
Soybean [Glycine max (L.) Merrill] is a major world crop. Total world hectares
planted increased from 47.1 million in 1977-78 to 57.7 million in 1989-90 (2,3). This
22.5% increase was due primarily to demand for oil and meal products, and to a small
extent the whole bean product (75). In the United States, hectarage rose to a peak of
28.6 million hectares in 1978-79 with a steady decline to 24.0 million hectares in 1989-
90. This 16% decrease in soybean hectarage was primarily due to reduced plantings in
the southeastern United States (3).
The reduction in hectares planted to soybean in the southeastern United States is
attributable to the low market value of soybean. Another reason for the reduction is the
inability of farmers, in general, to effectively produce a profitable crop in the presence
of yield-reducing factors. These factors include fungal diseases, viral diseases, bacterial
diseases, nematodes, insects, and drought. In 1982, Mississippi suffered an estimated
$50 million loss in revenue from insects alone, of which $40 million could be attributed
solely to the soybean looper, Pseudoplusia includens Walker (15). Similar losses also
occurred in states neighboring Mississippi as a result of insect infestations (15,16). Pest
problems which can be costly to control, and a 10-year (1981-1990) average price of
$6.19 per bu (3), make it increasingly difficult for a grower to produce a profitable crop.
1

2
Several techniques are available for control of diseases, nematodes, and insects
in soybean. Metcalf and Luckmann (50) categorized these techniques into seven major
methods: cultural, mechanical, physical, biological, chemical, genetic, and regulatory.
All methods of control have merit and should be utilized in integrated pest management
programs (14,20,58,50). Although economics is the major limiting factor in achieving
control of any pest of soybean, human safety and environmental impact must be
considered when choosing an approach to control a particular pest. Before making a
decision on what approach to take, a grower should answer the questions: ’Will the end
results justify the cost of control?’ and ’Is the control measure safe?’.
The soybean looper has become increasingly resistant to available insecticides
(10,18,48,58,60). For control of the soybean looper and other insect pests, a biological
approach should yield the best results. There are two biological approaches to insect
control: host plant resistance and the use of biocontrol agents (14,58). The latter
involves the identification and exploitation of natural enemies and pathogens. This
method of control usually is limited to a specific insect pest rather than the broad
spectrum of insects that damage soybean. As with conventional insecticides, resistance
to microbial insecticides has been observed (58).
Host plant resistance offers an effective method of control which is built into the
plant. It has relatively long stability, is compatible with other control tactics, is
environmentally safe (58), and development and subsequent release of a resistant cultivar
is far more socially acceptable (82). Lambert and Kilen (47) found that direct selection
for resistance in soybean to one species of insect resulted in the indirect selection for

3
resistance to certain other species. The development of an insect-resistant soybean
cultivar takes approximately the same amount of time as is required to develop a new,
effective insecticide, which can be as short as 10 to 12 years or as long as 20 years. The
insect-resistant cultivars ’Lamar’ (28) and ’Crockett’ (5) each took approximately 20
years to develop.
Initially, Van Duyn et al. (83) identified the three plant introductions PI 171451,
PI 227687, and PI 229358 from the USDA germplasm collection as having the highest
level of resistance to the Mexican bean beetle (Epilachna varivestis Mulsant). Schillinger
(69) identified additional sources of resistance to specific insects and described the
mechanism of resistance that each exhibits. Hatchett et al. (30) determined that breeding
lines identified as possessing resistance to multiple species of insects could serve as basic
germplasm for development of cultivars resistant to insects. To date, two soybean
cultivars and two soybean germplasm lines with resistance to insects have been registered
with the Crop Science Society of America (5,11,28,29).
Although there has been success in developing insect-resistant germplasm lines
and cultivars, there has been very little research conducted on the inheritance of insect
resistance in soybean (37,54,72). Of the three inheritance studies, two (54,72) were
conducted on the resistance to Mexican bean beetle and one (37) on the resistance to the
velvetbean caterpillar (Anticarsia gemmatalis Hubner). Since the soybean looper is a
major pest to soybean in the southeastern United States and is difficult to control with
insecticides, the inheritance of resistance to the soybean looper would aid breeders in the
development of resistant cultivars.

4
The objectives of this study were: (/) to evaluate three methods for rating
defoliation of soybean by soybean looper and (/'/') to determine the inheritance of
resistance in soybean to the soybean looper.

CHAPTER 2
LITERATURE REVIEW
Although undocumented, plants that are resistant to insects have survived and
propagated based on the processes of adaptation and natural selection. These ’survivors’
were in turn selected by farmers to plant future crops (74). This practice of selection by
primitive farmers could be considered the beginning of the development of insect-
resistant cultivars.
According to Kogan (38), Pedigo (58), and Smith (74), in their brief histories of
insect-resistant plants, the first development and subsequent cultivation of insect-resistant
cultivars occurred in the late eighteenth and early nineteenth centuries. In 1792, J.N.
Havens identified the wheat (Triticum aestivium L. em. Thell.) cultivar ’Underhill’ as
being resistant to the Hessian fly (Mayetiola destructor Say). In the 1830’s, G. Lindley
recommended the cultivation of apple (Malus pumila Miller) cultivars resistant to the
woolly apple aphid (Eriosoma lanigerum Hausman). In the mid-nineteenth century, the
French wine industry was saved by grafting the highly susceptible scions of the French
varieties to the rootstocks of American grapes (Vitis L. spp.), which were resistant to the
grape phylloxera (Phylloxera vittifolae Fitch).
In the United States, R.H. Painter is considered to be the founder and pioneer of
the modem era of research on insect-resistant plants (58,74). In his book, Insect
5

6
Resistance in Crop Plants, which was the first book written on the topic, he described
mechanisms of resistance in plants and factors that affect them. Based on field
observations, he separated the mechanisms of resistance in plants into three distinct, but
interactive classes: nonpreference, later known as antixenosis (38), (for oviposition, food,
or shelter), antibiosis (adverse effect on the biology of the insect), and tolerance (ability
to withstand, repair, or recover from infestation) (57). Developing a cultivar that
exhibits a high level of all three mechanisms of resistance would be optimum. This is
seldom feasible, therefore developing cultivars with the antibiosis mechanism of
resistance is the most frequent objective of plant breeders. Although nonpreference and
antibiosis differ in their mechanisms of resistance, both involve the interaction of the
biology of the plant with the biology of the insect (58,74). There is often an overlapping
of the nonpreference and antibiosis classes which makes it difficult to determine the exact
mechanism of resistance exhibited in a plant to a particular species of insect (74).
Although it is advantageous if a breeder knows which mechanism of resistance
is being expressed, knowledge of the mechanism is not essential to make advances in the
development of resistant germplasm. In fact, resistant cultivars and germplasm have
been developed for years with little concern for the underlying mechanism of resistance.
It has only been in the past two decades that mechanisms of resistance have been
identified in crop plants (39).
In soybean, Van Duyn et al. (83) identified three genotypes (PI 171451, PI
227687, and PI 229358) from the USDA germplasm collection of maturity groups VII
and VIII as resistant to the Mexican bean beetle. These three Pi’s have been used

7
extensively as sources of resistant germplasm and as standard checks in screening
additional germplasm for resistance. At the World Soybean Research Conference in
1975, Schillinger (69) reported additional sources of resistance to the Mexican bean
beetle. Gary et al. (22) screened 1108 lines from maturity groups VI, VII, VIII, and IX
of the USDA germplasm collection against velvetbean caterpillar initially, and
subsequently velvetbean caterpillar, soybean looper, corn earworm {Heliothis zea
Boddie), tobacco budworm (H. virescens Fabricius), and beet army worm (Spodoptera
exigua Hubner). Based on the results of the screenings against these five species, PI
209837 and FC 31592 were identified as having levels of resistance that would be
suitable for developing resistant cultivars. The germplasm collection of 473 Pi’s of wild
soybean, Glycine soja Sieb. & Zucc., was screened against soybean looper, velvetbean
caterpillar, beet army worm, and com earworm to identify additional sources of resistance
equal to, or greater than, Pi’s 229358 and 171451. McKenna et al. (52) identified three
lines (Pi’s 464935-1, 366119, and 407301) from this study as having less defoliation than
the two standards, thus suggesting that there are additional genes for insect resistance.
In 1988, Kraemer et al. (42) identified several accessions of the USDA germplasm
collection as being resistant to the Mexican bean beetle. They were from maturity
groups VI (FC 31665, and Pi’s 379621, 416925, and 416937) and VIII (Pi’s 417061 and
417136).
Upon identification of a genotype that is resistant to a certain insect species, the
genotype usually is screened against other insect species. Resistance to more than one
species will enhance its usefulness in a breeding program. Hatchett et al. (30) conducted

8
a study utilizing breeding lines that derive their insect resistance from PI 229358 to
determine if selection for resistance to one or two insect species would result in the
indirect selection for resistance to other troublesome species. They found that no single
insect species could be utilized to select for resistance to multiple species of insects
among adapted genotypes. In a similar study, Lambert and Kilen (47) determined that
when PI 229358 was used as the donor parent, direct selection for resistance to one
insect species resulted in indirect selection for resistance to other foliar-feeding species.
Research to develop soybean genotypes resistant to insects is usually conducted
with species that are indigenous to a particular area (19,25,35,44,54,64,80,83). Lambert
and Kilen (45) evaluated PI 229358, PI 227687, and PI 171451 against five insect species
to determine their relative levels of resistance. Insect species tested were velvetbean
caterpillar, soybean looper, com earworm, tobacco bud worm, and beet army worm.
Based upon the responses to these five species, they found that the mechanism for
resistance in PI 171451 was nonpreference, whereas that of PI 229358 or PI 227687 was
antibiosis. These same three Pi’s were evaluated against four important soybean
defoliating species (beet army worm, Porthesia taiwania Shiraki, Anómala cupripes Bates,
and Orgyia species) to determine their usefulness in developing breeding lines resistant
to insects indigenous to Taiwan. No single accessions had a consistently high level of
resistance to all of the insect species. PI 227687 demonstrated the highest level of
resistance to the most prevalent species, the beet armyworm, and therefore would be
suitable to initiate a breeding program (80).

9
With the sources of resistance to various species of insects identified, researchers
have channelled their energy into determining the basis of the mechanism of resistance
in certain soybean accessions. Since the mid-1970’s, researchers have investigated
various constituents of soybean to identify the underlying biochemical compound that
confers resistance to insects (9,53,73,81). If an association could be established between
a particular compound or class of compounds and resistance, then a new tool could be
developed to accelerate the selection process by breeders.
In studies with PI 227687, PI 229358, and two susceptible cultivars, Tester (81)
reported that resistant Pi’s had lower total nitrogen, more soluble carbohydrates, and
more total sterols at flowering and pod-fill than susceptible cultivars. Through a series
of grafting experiments, Lambert and Kilen (46) verified that constituents contributing
to resistance in PI 227687 and PI 229358 were formed in the leaves and translocated
throughout the plant. Smith (73) determined that the resistant factors in PI 227687
appeared to be chemical. He also found that resistance to the soybean looper in PI
227687 was an antibiotic effect due to the combined effects of a feeding deterrent and
a growth inhibitor. Chiang et al. (9) suggested that resistance to Mexican bean beetle
feeding in PI 227687 may be attributed to phenylpropanoid metabolites. In their
inorganic nutrient analysis of leaf tissue studies with soybean resistant to the Mexican
bean beetle, Mebrahtu et al. (53) found that the content of calcium, phosphorus, and
potassium was negatively correlated with pupal weight gain (r = -0.23**, r = -0.26**,
and r = -0.20*; p < 0.01 and p < 0.05, respectively).

10
In 1986, Kogan (39) published a treatise on the role of biochemicals in plant
resistance to insects. He examined the defensive patterns of three families: Solanaceae,
Cruciferae, and Leguminosae (specifically the common bean and soybean) and published,
as quoted below, a summation of all previous research findings that elucidated the
chemical bases of resistance in PI 171451, PI 227687, and PI 229358.
1. Foliar total nitrogen, soluble carbohydrates, organic acids
and sterols of both resistant and susceptible soybean vary
with stage of growth (Tester, 1977; Grunwald and Kogan,
1981).
2. Total nitrogen was generally lower and, at the end of the
season, soluble carbohydrates and sterols were higher in
the resistant lines; organic acids fluctuated without
correlation with resistance (Testor, 1977).
3. Sterol profiles were similar in both susceptible and resistant
lines, both in early and in late maturing soybeans
(Grunwald and Kogan, 1981).
4. Pinitol is a dominant cyclitol in soybean and it represents
an average of 60% of the total 80%-ethanol-soluble
carbohydrate fraction of soybean (Phillips et al., 1982).
5. Pinitol was found in large amounts in PI 229358, and in
smaller concentrations in the susceptible cultivar ’Davis’
(@ 1% dry weight). Added to artificial media at 0.8%,
pinitol reduced weight gain by corn earworm larvae by
about 63%. These findings suggest that pinitol is an
antibiotic factor in soybean but with no apparent antixenotic
property (Dreyer et al., 1979). As the reported results
could not be replicated elsewhere, this claim has been
recently disputed (Gardner et al., 1984).
6. Systematic fractionation of Pi’s 227687 (Smith and Fischer,
1983) and 229358 (Binder and Waiss, 1984) yielded
fractions with intermediate and with polar solvents that
caused slower developmental rates and high mortality in
soybean looper [Pseudoplusia includens (Walker)] and
Heliothis zea Boddie, respectively.

11
7. Seven phenolic acids were detected in PI 171451 and in cv.
’Forrest’; gallic, protocatechuic, p-hydroxybenzoic,
vanillic, caffeic, p-coumaric, and ferulic. The potentially
antibiotic caffeic and ferulic acids occurred at higher
concentrations in the resistant PI than in the susceptible
cultivar. In both types of plants, concentrations were
higher in injured than in uninjured tissue (Hardin, 1979).
8. Isoflavonoid phytoalexins in soybean cotyledons are potent
feeding deterrents for the Mexican bean beetle (Hart et al.,
1983). These phytoalexins are biochemically related to the
phenolics mentioned above, through the shikimic pathway.
(39, p.516-518).
Although a great deal of knowledge has been gained, there still remain many unanswered
questions as to what components comprise the various mechanisms of resistance.
A moderate amount of research has been conducted in the area of resistance in
soybean to insects, and much progress has been made in cultivar development.
However, there has been very little information published on the inheritance of resistance
to insects in soybean. In their work with the Mexican bean beetle, Sisson et al. (72)
hypothesized that resistance was inherited quantitatively and that the number of major
genes involved was small, perhaps two or three major genes. In further studies with the
Mexican bean beetle, Mebrahtu et al. (54) concluded that resistance appeared to be
primarily an expression of additive genetic variance. Kilen et al. (36), in their studies
with the soybean looper, suggested that susceptibility was partially dominant with a few
major genes conditioning resistance. In their studies with PI 171451, PI 227687, and PI
229358, and the F, plants derived from intercrosses among the three Pi’s, Lambert and
Kilen (45) suggested that genetic resistance in PI 171451 differed from that operating in
the other two lines. This was further supported by their work with the Pi’s and the F3

12
lines derived from the intercrosses among the Pi’s. They hypothesized that each PI
contained a resistance gene that differed from those of the other lines. To design a better
breeding program to develop cultivars resistant to the soybean looper utilizing breeding
lines that derive their insect resistance from PI 229358, it would be advantageous if the
mode of inheritance of this trait were known.

CHAPTER 3
EVALUATION OF RATING METHODS
Introduction
Throughout the past 25 years there have been advances in the identification and
development of soybean breeding lines resistant to phytophagous insects, especially
Lepidopterous species and the Mexican bean beetle (5,11,22,28,29,30,42,52,79,83).
Despite these advances, little effort has been devoted to improving the methods of
evaluating the amount of feeding injury of insects. Most researchers utilize a rating
system that estimates the level of resistance among cultivars, but little or no effort has
been made to determine whether or not the rating scales used were truly reflecting the
defoliation caused by insects. Hence, the objective of this study was to evaluate the
effectiveness of the widely used visual, or relative, assessment of insect defoliation as
compared to a quantitative, or absolute, assessment.
Literature Review
Plant pathologists have faced problems in the assessment of disease which is
dependent upon the causal organism and the host plant species. Many different methods
of disease assessment are available to plant pathologists (7). Shokes et al. (71) tested the
reliability of disease assessment procedures with late leafspot [Cercosporidium
personatum (Berk, and Curt.) Deighton] in peanut (Arachis hypogaea L.) and concluded
13

14
that plant pathologists would benefit if a standard rating system for foliar diseases of
peanut could be adopted as has been done in other crops. In a special report, Berger (4)
stated that a disease assessment procedure should have the following basic requirements:
easy to learn, quick and efficient to use, applicable across a broad spectrum of
conditions, and be accurate, precise, and reproducible. He concluded that the greater the
emphasis placed on the assessment method, the higher the quality of information obtained
by the method.
In 1980, Kogan and Tumipseed (41) described the various methods of quantitative
and qualitative assessments of defoliation that are currently used by researchers working
with phytophagous insects. These methods are: (/) photoelectric or electronic (measure
actual leaf area), (ii) gravimetric (regressing known leaf areas on the corresponding fresh
or dry weight of these areas, (iii) volumetric (volume displacement cylinder is used to
measure biomass), (iv) planimetric (mechanical measurement of area), (v) geometric
(estimates total nondefoliated leaf area so that percent defoliation can be calculated when
used in conjunction with a leaf area meter), and (v/) visual estimates (provide rapid
estimate of defoliation). Although descriptions of the methods were detailed by Kogan
and Tumipseed, no statistical comparisons were made among methods.
Some authors (58,78) refer to qualitative assessments as relative estimates and
quantitative assessments as absolute estimates. Relative estimates do not translate directly
into the amount of defoliation, whereas absolute estimates allow for direct translation into
the amount of defoliation caused by phytophagous insects. The major drawback of
absolute estimates is that they are usually time consuming and costly to make.

15
Conversely, relative estimates can be timely and are inexpensive which is why these
types of estimates are more widely used by researchers (74).
Funderburk et al. (19) determined that plot size or shape had little effect on the
relative rankings of soybean cultivars when measuring resistance to the velvetbean
caterpillar in terms of percentage of defoliation based on visual estimates. However,
they found that the mean and precision of the estimates were affected by plot size and
shape. The objective of this study was to compare two relative methods to one absolute
method for assessing defoliation caused by the soybean looper.
Materials and Methods
The three rating methods evaluated in 1990 and 1991 were the whole plant visual
rating (WP), the partitioned plant average visual rating (AV), and measured average leaf
area of the partitioned plant (AL). WP values were based on a l-to-10 scale divided into
10% increments with a score of 1 signifying defoliation of 0 to 10% and a score of 10
signifying defoliation of 91 to 100%. Although other scales, such as a logarithmic scale
(33), may be better at estimating defoliation, the l-to-10 scale was chosen since it is most
commonly used by participants in the Regional Host Plant Resistance Test which is
sponsored annually by the Southern Regional Information Exchange Group (SRIEG - 32).
With the AV and AL methods, the canopies were visually partitioned into three sections
(top, middle, bottom) enabling a more detailed assessment of soybean looper defoliation
on each plant. The partitioning of the plants was based on a procedure that was
developed by Plaut and Berger (59) that was cited by Shokes et al. (71). For AV

16
estimates, each section of the plant was scored using the same scale as the WP with the
three scores averaged to obtain the AV.
In 1990, a LI-COR LI-3000 (Li-Cor, Inc., Lincoln, NE 68504) portable leaf area
meter was used to measure the middle leaflet of a representative trifoliate from each of
the sections and the three measurements were averaged to obtain the AL. In 1990 the
AL method only allowed an indirect comparison to the other two methods, therefore the
AL was modified in 1991 to permit a direct comparison with AV and WP. In the 1991
AL method, a representative trifoliolate leaf from each section of the partitioned plant
was removed and taken into a laboratory to determine the area of each sample. The
portable leaf area meter was used to determine the remaining portion of each defoliated
leaflet (DEF) of each trifoliolate. Following the procedure described by Kogan and
Tumipseed (41), each leaflet of the trifoliolates was then placed on a photocopy machine
and a copy was made. Examples of leaflet photocopies are shown in Figure 1. These
silhouettes were then cut out to represent the trifoliolates in a nondefoliated state and the
area (NONDEF) for each was determined as mentioned above. The two values obtained
for each trifoliolate were used in the following formula to determine the percent
defoliation (% def): (NONDEF - DEF)/NONDEF * 100 = % def. The percent
defoliation for each of the three trifoliolates from each plant was then converted to the
same l-to-10 scale used by the other two methods and the values averaged, allowing
direct comparisons to be made among the three methods.
The plants used in the evaluation of the three methods were part of an inheritance
study that was conducted in field cages (43) in 1990 and 1991 at the Mississippi State

17
Figure 1. Examples of photocopied leaflets used in determining the % defoliation
in the AV (leaf area of partitioned plant) method.

18
University Delta Research and Extension Center, Stoneville, Mississippi. In the first
year of the study, seed were planted on 25 May 1990. In 1991, seed were planted on
25 May and 29 May in cage 1, and on 31 May in cage 2. Rainfall on 25 May prevented
completion of planting on the same day in cage 1.
To compare methods in 1990, ten soybean plants were chosen randomly from
each of 20 backcross families for a total of 200 plants. Fifty plants in each of two cages
were randomly chosen using PROC PLAN of PC-SAS (68) for use in the evaluation of
the three methods in 1991. These plants were chosen prior to release of soybean looper
moths in both years.
The 4000 three-day-old soybean looper moths were released on 25 June 1990
when plants were at the V5 to V6 stage of plant growth (17). Larvae from hatched eggs
were allowed to feed freely for approximately 12 days, at which time ’Centennial’ (27),
the susceptible check, was estimated to have a rating of nine according to the WP
method. The area within the cage was then sprayed with methomyl [S-methyl-N-((methyl
carbamoyl) oxy)-thioacetimidate] at 0.5 kg ha'1 to kill the larvae. This allowed
defoliation ratings to be made without further feeding. The day following spraying, each
plant was rated by all three methods.
Approximately 4000 three-day-old soybean looper moths were released into each
cage on 8 July 1991. At the time of release, plants in cage 1 were V7 to V8 and in cage
2 they were V6 to V7. The larvae from the hatched eggs were allowed to feed freely
until the susceptible check, Centennial, received a rating of ten by the WP method, 16
days after release of the moths. To cease feeding at this stage of plant growth, R1 to R2

19
in cage 1 and beginning R1 in cage 2, the plants in both cages were sprayed with
methomyl at 0.5 kg ha1 to kill the larvae. The following day each of the 100 plants was
rated using the WP and AV methods and samples were collected for the AL.
Data derived from each rating method for the 200 plants in 1990 and the 100
plants in 1991 were subjected to analytical procedures in PC-SAS (68). A correlation
analysis was run to determine the correlation coefficient (r) between WP and AL, AV
and WP, and AL and WP rating methods each year. A paired difference comparison
(76) was made between WP and AV methods in 1990 and between WP and AL, WP and
AV, and AL and AV methods in 1991. McNemar’s test for concordance (77) was used
to test for significant changes between WP and AL in 1990 and between WP and AL,
WP and AV, and AL and AV methods in 1991. This analysis was conducted to
determine if the methods differed in their estimates of defoliation. Precision, known as
the relative variation (RV), of the two relative methods for both years and the absolute
method in 1991 were calculated for each as follows:
RV = (SE/JC)100
where SE is the standard error of the mean and X is the mean. This relationship,
described by Ruesink (66) and Pedigo (58), is considered useful in estimating precision
of a sampling method by entomologists.
Results and Discussion
The 1990 scores for each of the three rating methods by plant are presented in
Appendix A. Scores for AV and WP were based on a l-to-10 scale, with lower numbers

20
indicating less defoliation. The measurements for AL in 1990 are presented as cm2 of
remaining tissue. Individual plant data for each of the three methods in 1991 are
presented in Appendix B. Unlike 1990, 1991 data for all three methods were based on
a l-to-10 scale. Means and standard deviations for both years are also reported in
Appendices A and B.
Frequency distributions of the 1990 WP and AV rating methods are shown in
Figure 2. The AL method was omitted from this graph because it measured the
remaining undefoliated tissue in cm2, whereas the AV and WP methods were used to
estimate percent defoliation on a l-to-10 scale. Frequency distributions of the AV and
WP methods suggest that either the evaluator underestimated defoliation by the WP
method or overestimated defoliation by the AV method. Without an absolute estimate
of defoliation, no substantial conclusion can be made as to the reliability of AV or WP
to accurately assess defoliation.
The 1991 frequency distributions of all three methods for both cages combined
are presented in Figure 3. Due to the modifications in the AL method, it was possible
to make direct comparisons between AL and the visual rating methods. The evaluator
overestimated defoliation to obtain too few plants below 30% or <4 in the l-to-10 scale
when using either relative method, WP or AV, as compared to AL, the absolute method.
Also, the evaluator had a tendency to overestimate the amount of defoliation over 50%
or >5 in the l-to-10 scale when using either relative method as compared to AL. Only
in classes 4, 5, and 6 was the evaluator’s estimate of defoliation with the WP and AV
methods similar to that of the AL. The problem of overestimating might be avoided in

21
Number of PI ants
Defoliation Scores
Figure 2. 1990 frequency distribution of defoliation estimates by the whole plant and
partitioned plant average visual rating methods. Defoliation scores based on l-to-10
scale where 1 = 0 to 10% defoliation, 2 = 11 to 20%, 3 = 21 to 30%, 4 = 31 to 40%,
5 = 41 to 50%, 6 = 51 to 60%, 7 = 61 to 70%, 8 = 71 to 80%, 9 = 81 to 90%, 10
= 91 to 100%.

22
Number of Plants
Defoliation Scores
Figure 3. 1991 frequency distribution of defoliation estimates by the whole plant,
partitioned plant average visual, and measured average leaf area of the partitioned plant
rating methods. Defoliation scores based on l-to-10 scale where 1 = 0 to 10%
defoliation, 2 = 11 to 20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to 50%, 6 = 51
to 60%, 7 = 61 to 70%, 8 = 71 to 80%, 9 = 81 to 90%, 10 = 91 to 100%.

23
the future if a logarithmic rating scale, such as the Horsefall-Barratt scale to assess
disease (33), were used to estimate defoliation. A logarithmic scale may have advantages
for assessment of defoliation because the visual acuity of the human eye is more adapted
to the identification of small differences at each end of the scale.
The plant scores of the AV and the WP rating methods in 1990 were highly
correlated (r = 0.65, p < 0.001). AL measurements were in cm2 of nondefoliated leaf
area remaining, therefore not a measure of defoliation. As there was no absolute
measure of defoliation, it was not possible to make further comparisons with AV and
WP.
In 1991, the scores for cage 1 from the AV and WP methods were significantly
correlated with the AL method (r = 0.86, p < 0.001; and r = 0.88, p < 0.001,
respectively). The AV and WP methods correlated well with each other (r = 0.88, p
< 0.001). By squaring each of the correlation coefficients (r), both the WP vs. AV and
WP vs. AL relationships had r2 values of 0.77 meaning 77% of the variance in one rating
method can be explained by the variation in the other. The AV vs. AL correlation had
a r2 value of 0.74, indicating that 74% of the variance of the AV method can be
explained by the variation in the AL method (or vice versa).
Although the results from ratings on plants in cage 2 were similar, correlations
were not as high as those in cage 1. Scores from the AV and WP methods were
significantly correlated with the AL method (r = 0.75, p < 0.001; and r = 0.67, p <
0.001, respectively). The AV and WP methods were significantly correlated (r = 0.74,
p < 0.001) with each other. Calculated r2 values were: WP vs. AV r2 = 0.55, WP vs.

24
AL r2 = 0.49, and AV vs. AL r2 = 0.56. These values are lower than those found with
the cage 1 data. Differences may be attributed to differences in the plant growth at the
time of rating or to evaluator error. The cage 2 ratings were made in the rain which
hindered the evaluator’s ability to accurately estimate defoliation. It was considered
more prudent to make ratings in the rain than to attempt to disregard the ensuing
regrowth which was increasing rapidly during a prolonged rainy period.
When data for both cages were pooled for analysis, scores for the AV and WP
methods were significantly correlated with the AL method (r = 0.79, p < 0.001; and
r = 0.78, p < 0.001, respectively). The AV and WP methods correlated well with each
other (r = 0.80, p < 0.001) as in the cages treated separately. The following r2 values
were obtained: WP vs. AV r2 = 0.64, WP vs. AL r2 = 0.60, and AV vs. AL r2 = 0.62
thus approximately 62% (± 2%) of the variation in one rating method could be attributed
to the variation in the other rating method.
In a similar study, a relative method and an absolute method were compared using
their estimates of defoliation of field plots infested primarily (90% of total insect
population) with the velvetbean caterpillar (G.R. Bowers, 1992, personal
communication). A l-to-9 scale, with 1 signifying the least damage, was used as the
relative method. The absolute method followed a procedure described by Kogan and
Kuhlman (40) in which 20 leaflets each from the top, middle, and bottom thirds of plants
(for a total of 60) were sampled from plants located in the center two rows of a four-row
plot. These were then compared to diagrams of defoliated leaves to determine the
defoliation estimate of each sample. The two methods were significantly correlated, r

25
= 0.62, p < 0.001. Bower’s correlation coefficients are somewhat similar to the pooled
correlation coefficients between the relative and absolute methods in this 1991 cage
experiment.
Correlations from the 1991 experiment, whether examined individually by cage
or combined, were all significant at p < 0.001. The magnitude of the r2 values indicate
that the evaluator obtained similar relative rankings of defoliation regardless the method
used. Although the correlations between the AL method and the WP and AV methods
were highly significant, it was not possible to determine if the evaluator estimated the
same amount of defoliation with the WP and AV methods (relative methods) as with the
AL method (the absolute method). To compare the accuracy of the evaluator using the
two relative methods and the absolute method, additional analyses were undertaken.
With the PROC MEANS procedure of PC-SAS, a paired difference comparison
was made between the WP and AV methods in 1990. A t-test was performed to test the
equality of the two methods based on the differences between the two methods for each
plant in the study. The mean difference of -0.36 between the WP and AV methods was
highly significant, p < 0.001, meaning the WP method yielded lower average percent
defoliation than the AV method. However, without some measure of the actual or
absolute defoliation across the plants, there was no way to determine if the evaluator
could obtain an accurate assessment of the true defoliation of a plant with either method.
A paired difference comparison was also made among the three rating methods
in 1991. The WP and AV methods had a mean difference of -0.55, p < 0.001, which
was somewhat similar in magnitude to the mean difference calculated in the 1990 study.

26
The mean differences ± standard deviation involving the AL method were 1.48 ± 0.14
for WP - AL and 2.03 ± 0.12 for AV - AL, respectively. Both differences were
significant at the p < 0.001 level. Although the evaluator estimated a higher average
percent defoliation with both relative methods than with the absolute method, the estimate
obtained with the WP method was numerically more similar to the measured defoliation
obtained with the AL method.
With the relative methods, WP and AV, the evaluator estimated somewhat similar
amounts of defoliation across all plants when compared with each other, even though the
difference between the two methods was highly significant. However, in comparisons
with an absolute method, AL, it becomes apparent that the evaluator’s estimates of
defoliation with the WP method were closer to the actual than were the estimates of the
AV method. Although the AL method is accepted as the absolute measure of defoliation
for each plant, there is some evidence that it underestimated defoliation. Necrotic leaf
tissue due to insect injury was present. The leaf area meter could not distinguish viable
leaf tissue from necrotic tissue. The amount of necrotic tissue was not enough to
increase AL values to the level of WP value, however the best absolute estimate of
defoliation would have been to sample the entire plant (destructive sampling) and follow
the protocol of the AL method after cutting out necrotic tissue. In the 1991 experiment,
provisions were not made for destructive sampling; therefore it was not practiced.
McNemar’s test for concordance was used to determine if the categorization of
plants having less than or equal to 50% defoliation and of plants having greater than 50%
defoliation differed with the WP and AV methods. In 1990, 36 of the 198 of the plants

27
were classified with the WP method as having at most 50% defoliation while with the
AV method these same plants were rated as having greater than 50% defoliation (Table
1). This percentage (18%) of plants was significantly different from zero, X2 = 34.03,
p < 0.001, supporting the earlier contention (Figure 2) that either the WP method
underestimated defoliation or the AV method overestimated defoliation. Without the
absolute method for reference it was not possible to determine which method was more
accurate.
All three methods were compared in 1991 to determine whether significant
differences among the methods existed. The results from McNemar’s test for
concordance of the 1991 data are shown in Table 1. The two relative methods were
significantly different (X2 = 11.25, p < 0.001) with 20 of the 97 plants being classified
differently by WP than AV. The relative methods were both significantly different from
the absolute method (WP vs. AL X2 = 22.04, p < 0.001; and AV vs. AL X2 = 38.03,
p < 0.001) in the estimate of defoliation on a plant-by-plant basis. With both relative
methods, the evaluator had a tendency to overestimate defoliation as compared to the
absolute method. With the AV method 40 of 97 plants were classified as > 50%
defoliated while AL classified these same plants as < 50%, whereas with the WP
method only 24 of 97 plants were classified > 50% as compared to the AL method
classifying them as < 50%. This indicates that the evaluator was able to estimate
defoliation closer to the absolute method with the WP method as was seen in the paired
difference comparison.

28
Table 1. McNemar test for significant differences between the WP (whole plant visual) and
AV (partitioned plant visual) methods in 1990; and between the WP and AV, WP and AL
(leaf area of the partitioned plant), and AV and AL methods in 1991.
AV
WP
1 -5t
1990
> 5t
Totals
X2
1 - 5
159
0
159
> 5
36
3
39
Totals
195
3
198
34.03***
1991
WP
AV
1 -5
> 5
1 - 5
39
2
41
> 5
18
38
56
Totals
57
40
97
11.25***
WP
AL
1 - 5
> 5
1 - 5
57
24
81
> 5
0
16
16
Totals
57
40
97
22.04“*
AV
AL
1 - 5
> 5
1 - 5
41
40
81
> 5
0
16
16
Totals
41
56
97
38.03***
*** Significant at the 0.001 level of probability.
t Defoliation ratings based on a 1 -to-10 scale where 1 = 0 to 10% defoliation, 2 = 11 to
20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to 50%, 6 = 51 to 60%, 7 = 61 to 70%, 8
= 71 to 80%, 9 = 81 to 90%, 10 = 91 to 100%.

29
RV values for 1990 were 2.11 and 2.77 for WP and AV, respectively. According
to Pedigo (58), RV values less than 10 are desirable in research, but the closer the value
is to zero the more precise the method is at making estimates. Although the WP method
was numerically better than the AV method, the precision of both methods was
essentially equal and very good.
The RV values for each of the methods in 1991 were as follows: WP = 4.26, AV
= 3.49, and AL = 4.09. Based on the RV values, the AV method was numerically
better than either WP or AL. The AL method was expected to be the most precise since
it was the absolute measure of defoliation. However in this study, these values were
essentially equal and close to zero, thus the evaluator was fairly precise in making
estimates of defoliation with all three methods.
Summary and Conclusions
The objective of this study was to evaluate the effectiveness of the widely used
visual, or relative, assessment of insect defoliation as compared to a quantitative, or
absolute, assessment. In 1990 no absolute method comparisons were possible, however
the WP and AV methods were highly correlated (r = 0.65, p < 0.001). The evaluator
estimated a slightly lower amount of defoliation on the average with the WP method than
with the AV. The relative variation of these two relative methods were both less than
3.0 which according to Pedigo (58) is very desirable. Without an adequate absolute
method no determination could be made as to which relative method best reflected the
actual defoliation.

30
Adjustments in the method of calculating percent defoliation by the AL method
for 1991 allowed for direct evaluations of the two relative methods. All correlations
(WP vs. AV, WP vs. AL, and AV vs. AL) were significant at the p < 0.001 level,
whether the cage data were treated separately or pooled. The two relative estimates of
defoliation were significantly correlated with the absolute estimate of defoliation. These
results are similar to the findings of a another relative versus an absolute estimation of
defoliation conducted in 1985 by G.R. Bowers (G.R. Bowers, 1992, personal
communication).
Paired comparisons were made among the methods to test the equality of the
methods to rate defoliation. In 1991, the estimates of defoliation by all three methods
were significantly different. These findings were validated by McNemar’s test for
concordance. The WP defoliation estimates were more similar to the AL results, than
the AV ratings were to AL.
Based on the RV’s of each method, the precision of the AV and WP methods was
essentially equal in 1990. Similar results were obtained in 1991 with the AV, WP, and
AL methods. In both years the precision of the methods was essentially equal and well
below the value of 10 that Pedigo (58) states as being desirable. These results indicate
that either relative method is suitable for estimating defoliation.
The results of these studies demonstrate that relative estimates of defoliation can
be used effectively in genetic studies or routine screenings of advanced breeding lines.
Since the two relative methods yielded similar results, it would be advantageous to use
the most economical method.

CHAPTER 4
INHERITANCE OF RESISTANCE
TO SOYBEAN LOOPER IN SOYBEAN
Introduction
During the past three decades great strides have been made in the identification
and development of insect-resistant breeding lines of soybean (22,30,42,52,79,83). This
research has led to the release of several germplasm lines and cultivars (5,11,28,29) that
are resistant to phytophagous insects. Although advances have been made, very little
information has been published on the inheritance of resistance to foliar-feeding insects.
Sisson et al. (72) reported that resistance to the Mexican bean beetle was
quantitatively inherited with two or three major genes involved. In other inheritance
studies, Mebrahtu et al. (54) concluded that resistance to the Mexican bean beetle was
primarily an expression of additive genetic variance. Kilen et al. (36) suggested that
susceptibility to the soybean looper was partially dominant and a few major genes
conditioned resistance. Based on their work with PI 171451, PI 227687, and PI 229358
and the F, progeny derived from the intercrosses among the three Pi’s, Lambert and
Kilen (45) suggested that the genetic basis for insect resistance in PI 171451 differed
from that in the other two Pi’s. This was further supported by their work with the three
Pi’s and the F3 lines derived from intercrosses among the three Pi’s (37). They
hypothesized that each PI contained a distinct resistance gene.
31

32
To facilitate the design of breeding programs to develop cultivars resistant to the
soybean looper, it would be beneficial to understand more completely the mode of
inheritance of this trait. Therefore, the objective of this study was to determine the mode
of inheritance of resistance to the soybean looper in genotypes derived from crosses
involving PI 229358.
Materials and Methods
All aspects of this experiment were conducted at the Mississippi State
University, Delta Research and Extension Center at Stoneville, MS on a Bosket fine
sandy loam (Mollic Hapludalfs).
1988
To study the inheritance of resistance to soybean looper in soybean, the breeding
line D86-3429 was crossed with the cultivar Braxton (32). D86-3429 has white flowers
(w7), gray pubescence (/), sensitivity to metribuzin [4-amino-6-(l,l-dimethylethyl)-3-
(methylthio)-l,2,4-triazin-5,(4H)-one] (hm), resistance to Phytophthora rot (induced by
Phytophthora megasperma Drechs. f. sp. glycinea T. Kuan & D.C. Erwin) (Rpsl-c), and
resistance to soybean looper. The donor of insect resistance was the germplasm line
D75-10169 (29), which derived its resistance from PI 229358 (83). The male parent,
Braxton, has purple flowers (W7), tawny pubescence (7), tolerance to metribuzin {Hm),
susceptibility to Phytophthora rot (rps), and susceptibility to soybean looper. Both
parents were of maturity group VII. Hand pollinations (1,60) made in the field produced
seven seeds of the cross D86-3429 x Braxton.

33
1989
The seven F, seeds produced in 1988 were space-planted 30.5 cm apart in 91.5-
cm rows in the F, nursery. Several rows of each parent, D86-3429 and Braxton, were
planted at a rate of 14 seeds per meter for crossing purposes. Four F, plants emerged
and these plants were used as the pollen parents for two backcrosses, D86-3429 x
F,(D86-3429 x Braxton) (BC,) and Braxton x F,(D86-3429 x Braxton) (BC2). The
original cross, D86-3429 x Braxton, and its reciprocal were also made to provide
additional F, seed for the following year. Approximately 60 hand pollinations were made
for the two backcrosses and the two crosses with the expectation that an adequate number
of seed would be produced. The two backcrosses, BC, and BC2, produced fourteen and
six seed, respectively. At maturity, the seed from the four F, plants were harvested to
provide F2 seed for use in 1990. In addition, seed from both parents were harvested for
use in 1990. Seven seeds were produced from the cross D86-3429 x Braxton and five
seeds were produced from the reciprocal cross. The seed of the backcrosses, BC, and
BC2, were scarified, treated with metalaxyl [N-(2,6-dimethylphenyl)-N-
(methoxyacetyl)alanine methyl ester] to optimize rapid emergence of healthy plants, and
planted one seed per pot in 15 L pots filled with soil treated with Bradyrhizobium
japonicum inoculum in the greenhouse during the winter of 1989-90. Plants were
allowed to self-pollinate, thus providing seed of the BS, (the selfed generation of BC,)
and BS2 (the selfed generation of BC2) generations. At maturity the plants were
harvested and threshed with an ALMACO (Allen Machine Co., Nevada, IA 50201)
single-plant thresher.

34
1990
The inheritance study was conducted in a screened field cage 2.4 m high X 19.2
m wide x 32.0 m long (43). A schematic of the field cage design is shown in Figure
4. In mid-April the soil was disked twice in opposite directions with a disk harrow. On
23 May the soil was tilled with a spring-tooth harrow to smooth the surface. This was
followed with a preplant-incorporated herbicide treatment of clomazone 2-(2-
chlorophenyl) methyl-4,4-dimethyl-3-isooxazolidinone and trifluralin [2,6-dinitro-N,N-
dipropyl-4-(trifluoromethyl)benzenamine] at 1.12 kg ha"1 and 0.84 kg ha'1, respectively,
on 24 May. The area within the cage was marked with rows spaced 68.5 cm apart to
be used as a guide for planting with push planters.
The seed for the study were planted on 25 May. A susceptible cultivar,
’Centennial’ (27), was planted at 27 seed per meter of row in border rows as well as a
1 m section at the ends of each row of experimental material. Seed to produce
experimental plants were scarified, treated with metalaxyl, and space planted 25 cm
apart. The parents, D86-3429 and Braxton, were planted in three 12.7 m rows for a
total of 150 plants each. Also, 50 seed from each of the fourteen BC, and six BC2 plants
were planted in 12.7 m rows. Following the backcross progeny, five Fj seed each of the
cross D86-3429 x Braxton and its reciprocal were planted. These were followed by the
360 F2 seed from the F, (D86-3429 x Braxton) plants grown in 1989. The insect-
resistant germplasm lines D75-10169 (29) and PI 229358 (83) were planted in the middle
and at the end of the experimental material for a total of 70 plants each.

Figure 4. Schematic drawing of field cages used in the 1990 and 1991 experiment.

36
Normal production practices were used to maintain healthy plants. Soil was
cultivated on 7 and 20 June to control weeds, and bentazon [3-(l-methylethyl)-(lH)-
2,l,3-benzothiadiazin-4(3H)-one2,2-dioxide] was applied at 0.84 kg ha'1 on 19 June with
a hand held pneumatic sprayer. Supplemental irrigation was not required during this
time. A Saran cage top (43) was installed on 21 June as the final step in preparation for
the release of insects.
Approximately 4000 laboratory-reared (26) soybean looper moths were released
in the cage at 0830 h on 25 June. The absence of physical restrictions made all areas of
the cage equally accessible to the moths. Water was applied by furrow irrigation to
elevate the relative humidity within the cage to enhance egg laying. All plants were at
the V5 to V6 growth stage at the time of release. Larvae emerged from eggs after an
incubation period of approximately three days.
Larvae were allowed to feed until susceptible Centennial was determined to be 85
to 90 % defoliated which occurred on 6 July or 11 days after moths were released. At
that time, methomyl was applied at 0.5 kg ha1 with a backpack sprayer to kill the
loopers and stop further feeding. On 9 July, the Saran cage top was removed to facilitate
regrowth by plants that were to be harvested for use in 1991, and all plants in the study
were evaluated for defoliation. Each plant was assigned a score of l-to-10 (AV method)
based on the visual estimate of defoliation. The l-to-10 rating scale represented 10%
increments, with a score of 1 representing defoliation of 0 to 10% and a score of 10
representing defoliation of 91 to 100%.

37
Ratings for the two parents, D86-3429 and Braxton, and their F, and F2 progeny
were grouped by their defoliation scores and plotted on a graph. The distribution of the
parental and F2 populations were tested for normality following the procedure described
by Leonard et al. (49). Initially the defoliation data of these populations were
summarized in frequency distributions by using the upper class limits. The mean x,
variance s2, and standard deviation s of each population were calculated from the class
centers (i.e. 0.5, 1.5, 2.5, ... 9.5) using the following formulas:
x = n^O.5) + n2(l .5) + ... + n10(9.5)/(n, + n2 + ... + n10)
s2 = n^O.5 - x) + n^l.5 - x) + ... + ^(9.5 - x)/(n.. - 1)
s = (s2)1/2
where n„ was the number of plants in each of the respective classes and n.. was the grand
total. Following the calculation of these variables, the variable x was calculated for each
class in the following manner:
x = (x - upper class limit)/s
After determining the x for each class, a table of normal probability integrals was used
to determine the corresponding Z value of each class. The theoretical percentage of each
population was calculated by subtracting Z from 1.000 and multiplying by 100. The
theoretical number of each class was calculated as follows:
n.. x theoretical % = theoretical n

38
For each succeeding class, the cumulative percentage in all prior classes must be
subtracted to obtain the theoretical percentage to calculate the theoretical n of that class.
The observed distribution versus theoretical distribution was tested with a X2 test. This
was done to gain insight into the mode of inheritance of resistance to soybean looper and
to facilitate the design of the 1991 experiments. The information from these four
populations was used to determine what additional crosses, if any, would be required to
fulfill the objectives of this study.
The original cross, D86-3429 X Braxton, was repeated to generate seed to be
used as the F, generation in 1991. The two backcrosses, BC, and BC2, were repeated
for the same purpose as the original cross, as well as to provide seed for planting in a
winter nursery to generate the BS, and BS2 populations, the selfed generations of the
backcrosses. The bulk of the pollinations were made on plants within the cage. The
remainder were on plants located in the F2 nursery. Plants within the cage were irrigated
as needed throughout the remainder of the growing season to insure adequate seed set on
F, and F2 plants for use the following year.
Twelve seeds were produced from the D86-3429 x Braxton cross. The two
backcrosses, BC, and BC2, yielded 49 and 29 seeds, respectively. The 12 F, seeds were
combined with remnant D86-3429 x Braxton F, seed generated from the 1988 and 1989
crosses for use in 1991. Twelve seeds and 11 seeds, from BC, and BC2, respectively,
were scarified and treated with metalaxyl and sent to the winter nursery at Mayaguez,
Puerto Rico to produce the BS, and BS2 generations. The single plants from the various
generations in the cage were harvested and threshed with a single-plant thresher between

39
31 October and 2 November. The soil in the cage was subsoiled on 9 November in
preparation for the 1991 experiment.
1991
Two field cages were used to complete the inheritance study. Soil in each cage
was disked with a disk harrow and treated with trifluralin at 0.84 kg ha1, which was
incorporated with a spring-tooth harrow on 26 April. To control a broad spectrum of
weeds, a tank mix of clomazone at 1.12 kg ha'1 and imazaquin [2-(4,5-dihydro-4-methyl-
4-( 1 -methylethyl)-5-oxo-1 H-imidazol-2-yl)-3-quinolinecarboxylicacid] at 0.11 kg ha'1 was
applied and incorporated with a spring-tooth harrow on 13 May. On 25 May, soil was
reworked with a spring-tooth harrow and rows spaced 68.6 cm apart were marked to
serve as guides for the push planter.
Each row consisted of 10 plants spaced 25 cm apart except for the Fj, BC1; and
BC2 generations, which had only one plant per row. There were three rows, consisting
of 10 plants each, of the insect-resistant germplasm lines D75-10169 and PI 229358.
The planting order for each cage was generated by randomizing the 277 rows with PC-
SAS PROC PLAN (68). Each cage contained the entire complement of plants and was
treated as a replication over environments.
The border rows were planted to Centennial at 27 seed per meter in each cage.
The rows of experimental populations were centered 68.6 cm apart. Rainfall of 43 mm
halted planting on 25 May in cage 1 after one third of the rows were planted. The
remainder of cage 1 and all of cage 2 was lightly tilled with a spring-tooth harrow on 29
May. The remaining rows in cage 1 were planted on 29 May. Approximately 1 m of

40
Centennial was planted at the ends of all rows to serve as a border. The experimental
populations were surrounded by a susceptible cultivar.
On 31 May, soil in cage 2 was rolled with a water-filled roller pulled by a four-
wheel all-terrain vehicle. This operation smoothed and firmed the seed bed, which
decreased the rate of evaporation of moisture from the soil. The rows were then marked
as in cage 1. The susceptible cultivar Centennial was planted by the same method as in
cage 1. The entire area within cage 2 was planted on 31 May. Short sections of the
Centennial border rows were replanted on 14 June.
Soil in the cages was cultivated on 4, 14, 20, and 26 June with a self-propelled
garden tiller. Supplemental soil moisture was provided by furrow irrigation on 4 and 21
June to ensure that the plants would have sufficient vegetative growth at the time of
insect release. The Saran tops were put on the cages on 26 June as final preparation for
the release of insects.
Approximately 4000 laboratory-reared (26) soybean looper moths were released
in each cage at 0800 h on 8 July. A rainfall of 19 mm on 4 July provided sufficient soil
moisture to maintain an elevated relative humidity within the cages during the egg-laying
period. At the time of release, the plants were at the V7 to V8 stage of growth in cage
1, and V6 to V7 in cage 2 (17). Larvae emerged after an incubation period of
approximately three days.
Larvae were allowed to feed in each cage until 22 July, at which time Centennial
was approximately 95% defoliated. The cage tops were then removed to facilitate the
defoliation ratings and to promote regrowth of the plants. Both cages were sprayed with

41
methomyl at 0.5 kg ha'1 to kill larvae. This insured that additional feeding would not
occur once the ratings had begun. Cage 1 was rated on 23 and 24 July using the AV
rating method. This method used the same l-to-10 scale previously described for the WP
method in 1990. In the AV method the plants were visually partitioned into three
sections (top, middle, bottom). Each section was scored for defoliation and these three
scores were averaged to get the mean score for each plant. This method enabled the
evaluator to make a more detailed assessment of defoliation by the soybean looper on
each plant. Cage 2 was rated during a rainfall by the same method on 25 and 26 July.
All ratings in both cages were made by the same evaluator.
Generation means and variances were calculated for each of the nine generations
on a cage basis. These means and variances were employed in Mather’s scaling tests
(51) to determine the adequacy of an additive - dominant model and to test for epistasis.
The formulas for the different scaling tests were as follows:
A = 2xBC, - xP, - xFj VA = 4VxBC, + VxP, + VxF,
B = 2xBC2 - xP2 - xF, VB = 4VxBC2 + VxP2 + VxF,
C = 4xF2 - 2xF, - xP! - xP2 Vc = 16VxF2 + 4 VxF, + VxP, + VxP2
D = 8xF3 - 3xP, - 3xP2 - 2xF, VD = 64VxF3 + 9VxP, + 9VxP2 + 4VxF,
where x was the mean of the respective generations and Vx was the variance of the mean
of the same respective generations. The variance of the mean of each generation was
calculated as Vx = a1 In, where a2 was the variance of the generation and n was the
number of individuals observed in that generation.

42
To determine if each of the scaling tests had the expected value of zero, each was
tested in the following manner:
tA = A/(Va)*
h = B/(Vb)*
tc = C/(Vc)*
tD = D/(Vd)*
where (VA)‘A, (VH)'A, (VC)'A, and (Vn)'A were the standard errors of each tests. The t
values obtained were compared to a table of t distribution (76) to determine the level of
significance for each of the tests. The probability was found in the table of t using the
sum of the degrees of freedom of each generation in each test as the number of degrees
of freedom.
Since the preliminary results indicated that resistance to soybean looper is
inherited quantitatively, it was necessary to select an appropriate method of analysis that
would allow for a detailed study of the inheritance of resistance. After reviewing the
statistical methods available for the analysis of a quantitative trait, a generation means
analysis was selected. This analysis was chosen over other methods because of the
following advantages (23):
1. Errors are inherently smaller since means (first order
statistics) are used, rather than variances (second order
statistics);
2.
Smaller experiments allow the same degree of precision;

43
3. Additive (a), dominant (d), and epistatic (aa, ad, and dd)
effects are estimated using means rather than variances;
and
4. It is applicable to both cross- and self-pollinating crops.
In utilizing the generation means analysis, it is assumed that the parents are homozygous
and there is no linkage among genes influencing the trait being studied. If this
assumption is correct and it appears that the trait in question is not quantitatively
inherited, these generations can be used to analyze for Mendelian ratios.
Hallauer and Miranda (23) cited the following disadvantages to the generation
means analysis:
1. An estimate of heritability cannot be obtained; and
2. Genetic advances cannot be predicted because genetic
variances are not available.
Although these disadvantages are inherent to the generation means analysis, they are of
little consequence since additional calculations can be made to obtain the variances
necessary to estimate heritability.
Generation means analysis, as described by Hay man (31), has been used
extensively in other crops such as com (21,34,70), cotton (Gossypium hirsutum L.)
(55,56), and wheat (8). In most cases, Gamble’s (21) notation is used in conjunction
with Hayman’s (31) methodology. Generation means analysis apparently has not been
used previously in soybean to analyze quantitative traits. Because of the advantages of

44
this analysis, it was chosen to determine the mode of inheritance of resistance to soybean
looper.
Populations used were derived from the cross D86-3429 x Braxton (Figure 5).
A minimum of six populations are required in a generation means analysis with a six
parameter model, but this experiment was expanded to include nine populations. The
number of individuals comprising each generation are listed in Table 2. To determine
the inheritance of resistance to soybean looper in soybean according to Hayman’s (31)
model, programs were written in SAS to analyze the data by cage (see Appendix C) and
combined over cages where cages were treated as blocks (see Appendix D). Using
Hayman’s methodology and Gamble’s (21) notation, the models for the generation means
analysis were as follows:
p,
=
m
+
a
- d/2
+
aa
-
ad
+
dd/4
p2
=
m
-
a
- d/2
+
aa
+
ad
+
dd/4
F,
=
m
+ d/2
+
dd/4
f2
=
m
BC,
=
m
+
a/2
+
aa/4
bc2
=
m
a/2
+
aa/4
f3
=
m
- d/4
+
dd/16
BSj
=
m
+
a/2
- d/4
+
aa/4
-
ad/4
+
dd/16
bs2
=
m
-
a/2
- d/4
+
aa/4
+
ad/4
+
dd/16
where m = the overall mean, and a, d, aa, ad, and dd represent the additive, dominance,
additive x additive, additive x dominance, and dominance x dominance genetic effects,
respectively.
Each SAS program fitted two regression models which were set up using matrix
notation according to the procedures outlined by Jennings et al. (34). The first

45
D86-3429 (P,) x Braxton (P2)
P, x
F,
x P,
BC,
BC,
BS,
BS,
Figure 5. Mating scheme to derive populations required to estimate the genetic
effects of resistance to the soybean looper in soybean.

46
Table 2. Populations used for generation means analysis of the inheritance of resistance to
the soybean looper.
Population
Gen.
Row
Number
Plants
per Rowf
Total
Plants
D86-3429
Pi
1 - 11
10
110
Braxton
P2
12 - 22
10
110
D86-3429 x Braxton
F,
23 - 35
1
13
D86-3429 X Braxton
f2
36 - 85
1
500
D86-3429 x Braxton
f3
86 - 210
10
1250
D86-3429(2) X Braxton
BC,
211 - 224
1
14
D86-3429(2) X Braxton
BS,
225 - 249
10
250
Braxton(2) x D86-3429
bc2
250 - 255
1
6
Braxton(2) x D86-3429
bs2
256 - 271
10
160
D75-10169$
-
272 - 274
10
30
PI 229358$
-
275 - 277
10
30
fRows were 2.5 cm long except where there was only 1 plant/row.
^Insect-resistant germplasm lines used as checks.

47
regression model, defined as Model 1, consisted of the three parameters m, a, and d.
The second regression model, defined as Model 2, consisted of the epistatic effects, aa,
ad, and dd, in addition to the parameters in Model 1. Model 2 is used only if a
significant additive or dominant effect is detected and to determine if significant epistatic
effects exist that are contributing to the significance in Model 1. To adjust for the
unequal population sizes comprising each generation, the models were weighted using
reciprocals of the standard errors of the generation means as suggested by Rowe and
Alexander (65).
In addition, Powers’ partitioning method of genetic analysis (49,61) was
incorporated to test proposed genetic models. Initially the frequency distributions of the
Pl5 P2, F,, and F2 generations were tested for normality by testing the observed versus
theoretical distribution with a X2 test following the procedure previously outlined.
Frequency distributions (observed and theoretical) of the F2 population were converted
to percentages in each class in the following manner:
%(in each class) = n¡/n..
where n¡ was the number of plants in each class and n.. was the total number of plants
in the F2 population. After conversion, the observed distribution of the F2 population
was examined for modes. Finally, the F2 population of each cage was tested for
goodness of fit to a two-gene additive model. The proposed genetic ratio for the model
was as follows: 1:2:1:2:4:2:1:2:1. To test the model, the ratio was simplified to 1:14:1
by combining the defoliation classes in the two cages as follows: 2-4, 5-8, and 9-10 in

48
cage 1, and 2-3, 4-8, and 9-10 in cage 2. The end classes were defined by a distinct
transition between defoliation class in the distribution of the F2 population of each cage.
The homozygous populations D86-3429, Braxton, D75-10169, and PI 229358
were used to determine if there was a significant location effect within each cage. An
analysis of variance was performed on each population to obtain the row and plant within
row mean squares. Estimates for plant to plant variability within a row (a2p) and the
additional variability of plants in different rows (o2,) were obtained from the expected
values of the mean squares. Also, F-tests for significant row effects (Ho:o2r=0) were
performed on each of the homozygous generations within each cage.
An estimate for the number of genes (n) involved in resistance to soybean looper
was obtained for the progeny in each cage by the formula:
n = (xP, - xP2)2/8[(o2F2) - ((v/a2P1 + v2(j2P2-l-vJa2F1)/v2-l-v2+vi)]
In the formula, xPj and xP2, were the mean defoliation ratings for the parents, and
a2P1; a2P2, and j^F, were the variances of the respective generations. This formula is
a modification of the formula presented by Poehlman (60). In the modification, a2F, is
replaced with (v/a2P1 + v2a2P2+via2F1)/(vi + v2+vi). This modification was used in order
to better estimate the environmental variance, since a2Fb and o2P2 were essentially
estimates of environmental variation. The degrees of freedom (v;, v2, and v3) for the
associated generations were employed to lessen population size effects on the estimate
of environmental variance. This method of estimating genes assumes that the genes have
equal effects without significant dominant or epistatic effects.

49
Heritability of insect resistance, in the broad sense, was estimated for the progeny
in each cage from the formula:
H = (Vo/Vp) x 100%
where VG was the genetic variance and VP was the phenotypic variance. The genetic
variance was:
VG = VP - VE
where VE was the environmental variance. The VE was estimated by obtaining the
weighted average of the variances of the F, population and the populations of the parents
P, and P2:
(v; cFP, + v2o2P2+VjcFF^/O;+v2+vj)
In the formula the respective degrees of freedom v were used to weight the variances of
each population. The VP was estimated by utilizing the variance of the F2 population.
Results and Discussion
1990
The defoliation data of the P,, P2, F,, and F2 populations from the preliminary
cage study are presented in Table 3. Also presented in this table are the defoliation
ratings of PI 229358 and D75-10169, from which D86-3429 derived its resistance to
soybean looper. D86-3429 was visibly more resistant than the original source of

50
Table 3. Ratings of leaf feeding by soybean looper on soybean parents, their F, and F2
progeny, and the germplasm lines PI 229358 and D75-10169 in the field cage-1990.
Upper class limits of leaf feeding ratings!
n
1
2
3
4
5
6
7
8 9
X
D86-3429(P,)
4
60
64
11
2.6
139
Braxton(P2)
1
32
48
38
19 1
6.3
139
P, X P2(Fj)
5
3.0
5
P2 X P^Fj)
1
4
3.8
5
Pi X P2(F2)
1
33
78
154
39
6
2
1
3.7
314
PI 229358
21
25
1
1
3.6
48
D75-10169
4
34
18
3.3
56
fl = 0 to 10% defoliation, 2 = 11 to 20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to
50%, 6 = 51 to 60%, 7 = 61 to 70%, 8 = 71 to 80%, 9 = 81 to 90%, 10 = 91 to
100%. None of the plants received a rating of 10.

51
resistance, PI 229358. This suggests that additional genes for resistance might be
contributing to the increased level of resistance.
Resistance to soybean looper was interpreted as being quantitatively inherited
based on the defoliation data of the P1? P2, F,, and F2 populations (Table 3). Distribution
of the F2 population (n=314) was continuous, with a mean defoliation rating of 3.7
which was slightly less than the parental midpoint of 4.5. Although a high number of
individuals fell in the fourth rating class (31 to 40% defoliation), the F2 population
appeared to have a normal distribution. The P,, P2, and F2 population distributions were
tested for normality using the procedure outlined by Leonard et al. (49) (see Appendix
E). Both parental populations had low X2 values (Table 4) indicating they were normally
distributed. The frequency distribution of the F2 ratings failed the normal distribution
test (Table 4). Rating class 4 had too many plants, whereas classes 3 and 5 had too few
(Table 3 and Appendix E).
Although the defoliation ratings of the F2 population did not fit a normal curve
the distribution was continuous with only one mode. Therefore use of a method of data
analysis for quantitative traits seemed appropriate in order to define the mode of
inheritance of soybean looper resistance. A generation means analysis (31,34) was
chosen and additional crosses were made to supply the generations necessary for the
experiment.
1991
Larvae were allowed to feed until susceptible Centennial was 95% defoliated.
Plants were found in all rating classes except class one. All populations receiving larger

52
Table 4. Goodness of fit test for normalityf of soybean looper defoliation for
populations of D86-3429, Braxton, and D86-3429 x Braxton F2 populations in the 1990
cage study.
Population
Degrees of Freedom
X2
P
D86-3429
3
3.6628
0.50 - 0.25
Braxton
5
7.1877
0.50 - 0.25
f2
7
22.3204
< 0.005
f Test follows the procedure of Leonard et al. (49).

53
mean defoliation scores in 1991 (Table 5) than in 1990 (Table 3), when larvae were
killed when Centennial was only 85% defoliated.
In 1991 the AV rating method was substituted for the WP method used in 1990.
The AV method accounted for feeding preferences of soybean looper (24). It was
expected to yield a better estimate of defoliation than the WP method, because each score
of the AV method was the average of three observations per plant, whereas the WP
method was based on a single observation. The AL method was too time consuming to
be used in an experiment of this magnitude with the resources available.
Ratings of defoliation by soybean looper on the parents, P, and P2, and their Fb
F2, and F3 progenies grown in cages 1 and 2 are shown in Table 5. The F2 and F3
progenies, in both cages, exhibited a normal distribution for ratings of defoliation. In
both cages the F, population exhibited a bimodal distribution instead of the expected
single mode. All of the F, plants exhibited purple flowers and tawny pubescence, ruling
out the possibility that the bimodal distribution resulted from seifs in D86-3429.
The F, defoliation data distribution was expected to exhibit a narrow range and
single mode intermediate to the two parental modes. Instead the distribution had a wide
range and was bimodal in both cages (Table 5). Since the F! data did not fit expected
distributions, it was suspected that there was a location effect within each cage interfering
with the defoliation ratings. The four homozygous populations (D86-3429, Braxton,
D75-10169, and PI 229358) included in the study were tested to determine if there was
a significant location effect within each cage. A significant location effect was detected
for the populations of D86-3429, Braxton, and D75-10169 in each cage (Table 6).

54
Table 5. Ratings of leaf feeding by soybean looper on soybean parents, their F,, F2) F3, BC,,
BSl5 BC2, and BS2 progeny, and the germplasm lines PI 229358 and D75-10169 in the field
cages-1991.
Upper Class Limits of Leaf feeding ratings!
n
2
3
4
5
6
7
8
9
10
x$
Cage 1
D86-3429(P,)
3
30
44
19
2
3.5
98
Braxton(P2)
2
18
30
40
4
8.0
94
P. X P2(F,)
1
2
2
6
5.8
11
P. X P2(F2)
1
8
31
58
102
111
67
36
4
6.2
418
P. X P2(F3)
7
37
118
194
232
201
148
131
18
6.0
1086
P, x (P, x P2)(BCj)
5
4
1
1
1
2
4.4
14
P, x (P, x P2)(BSj)
2
34
64
61
19
8
6
4.2
194
P2 x (P, x P2)(BC2)
3
1
7.9
4
P2 x (P, x P2)(BS2)
1
10
13
52
45
9
7.9
130
PI 229358
1
1
7
11
1
5.1
21
D75-10169
2
7
10
7
2
3.6
28
Cage 2
D86-3429(P,)
8
32
43
14
3.3
97
Braxton(P2)
1
5
34
43
11
8.3
94
P. X P2(F,)
1
2
1
2
3
2
1
5.4
12
P. X P^F,)
2
17
43
58
82
120
93
30
5
6.1
450
P. X P2(F3)
28
103
189
238
273
187
114
21
6.2
1153
P, x (P, x P2)(BCl)
2
5
3
2
2
4.5
14
P, x (P, x P2)(BSj)
5
23
59
51
37
16
7
1
4.5
199
P2 x (P, x P2)(BC2)
2
9.3
2
P2 x (P, x P2)(BS2)
1
5
12
17
33
43
11
7.7
122
PI 229358
3
16
8
1
4.9
28
D75-10169
3
9
9
6
4.4
27
fBased on the mean of three observations per plant. 1 = 0 to 10% defoliation, 2 = 11 to
20%, 3 = 21 to 30%, 4 = 31 to 40%, 5 = 41 to 50%, 6 = 51 to 60%, 7 = 61 to 70%, 8 =
71 to 80%, 9 = 81 to 90%, 10 = 91 to 100%. None of the plants received a rating of 1.
^Derived from the mean defoliation of all plants without regard of the upper class limits.

55
Table 6. Estimates of variance components obtained from an analysis of the homogeneous
populations, D86-3429, Braxton, D75-10169, PI 229358, for a location effect due to
feeding1- by soybean looper in field cages in 1991.
Population
«V
crip + oV
a
0
•n
II
0
Cage 1
D86-3429 (P,)
0.31
0.72
12.55***
Braxton (P2)
0.25
0.84
20.78***
D75-10169 (BD)
0.29
1.50
40.29“*
PI 229358 (D)
0.58
0.96
5.37*
Cage 2
D86-3429 (P,)
0.37
0.54
5.03***
Braxton (P2)
0.69
0.80
2.35*
D75-10169 (BD)
0.42
0.94
12.03*“
PI 229358 (D)
0.39
0.41
1.41
*, **’ Significant at the 0.05 and 0.001 probability levels respectively.
f Based on defoliation ratings made with a 1 to 10 scale divided into 10% increments with
a 10 reflecting 91-100% damage.
* Variance of plants within same row.
§ Variance of plants among rows.

56
The location effect for the PI 229358 population was significant (p < 0.05, Table 6) in
cage 1, but was nonsignificant in cage 2. Although the population size of PI 229358 is
approximately the same for each cage, different randomization plans would produce
different results. It is apparent that the defoliation ratings were affected by location
within the cages.
The mean ratings for soybean looper defoliation and their standard errors for the
nine generations grown in each cage are presented in Table 7. In addition to the
observed mean ratings for defoliation, the predicted means determined from relationships
described by Mather and Jinks (51) are also presented. In all instances except the BC2
generation in cage 2, the predicted means were within ± one standard deviation of the
observed means in each cage. The observed mean for the BC2 generation in each cage
was larger than the predicted mean. This deviation from the predicted mean could be
associated with the small sample size available for defoliation ratings. Variances of the
generations from each cage are presented in Table 7, and are used later in Mather’s
scaling test (51) and to determine heritability.
The results of Mather’s scaling test (51) for the cross D86-3429 x Braxton are
shown in Table 8. According to Mather and Jinks (51), if the values of A, B, C, and
D do not differ from zero, then additive gene effects are indicated. The significance of
any of these scales indicates the presence of epistasis. In cage 1, nonsignificant values
for A, B, C, and D, indicated that an additive-dominance model was adequate. Cage 2
data yielded similar r-tests for A, C, and D, supporting the conclusions from cage 1.
The B scale test could not be applied to cage 2 since the BC2 population had a variance

57
Table 7. Means, standard errors, and variances of soybean looper defoliation ratings for
parental lines and seven populations of soybean arising from the cross D86-3429 x
Braxton grown in field cages at Stoneville, MS in 1991.
Population
Gen.
Number Actual
Mean
Predf
S.E4
o2
D86-3429
P.
98
Cage 1
3.49
0.084
0.685
Braxton
P2
94
7.97
-
0.092
0.797
Pi X P2
F,
11
5.76
5.73
0.427
2.002
Pi x P2
f2
418
6.19
5.75
0.072
2.160
Pi x P2
f3
1086
6.02
5.96
0.052
2.916
Pi x (Pi x P2)
BC,
14
4.36
4.63
0.524
3.837
Pi * (Pi x P2)
BS,
194
4.22
4.62
0.087
1.452
P2 x (Pi x P2)
bc2
4
7.92
6.87
0.394
0.620
P2 X (P, X P2)
BS2
130
7.89
6.86
0.090
1.055
D86-3429
Pi
97
Cage 2
3.34
0.074
0.528
Braxton
P2
94
8.28
-
0.092
0.788
Pi x P2
F,
12
5.42
5.81
0.617
4.568
Pi X P2
f2
450
6.13
5.62
0.074
2.490
Pi x P2
f3
1153
6.18
5.97
0.046
2.478
Pi * (Pi * P2)
BC,
14
4.45
4.38
0.376
1.976
Pi x (Pi x P2)
BS,
199
4.52
4.48
0.093
1.708
P2 X (Pj x P2)
bc2
2
9.33
6.85
0.000
0.000
P2 X (Pj x P2)
bs2
122
7.69
6.95
0.117
1.655
•¡•Based on relationships described by Mather and Jinks, 1971.
^Standard error

58
Table 8. Mather’s scaling test applied to the cross D86-3429 x Braxton to test adequacy of
additive-dominance model for resistance to soybean looper.
Test
Cage 1
Cage 2
ÍA
-0.125
0.039
0.918
-
0.268
0.267
0.157
0.273

59
of zero. Since C was not significantly different from zero in either cage (Table 8), this
indicated that generation means depended only on the additive gene effects and no
epistasis was present.
The results of the generation means analysis of each cage are reported in Table
9. Table 10 reports the results of the combined analysis where the cage effects are
treated as blocks. Significant negative additive effects were detected with Model 1,
which measures only additive and dominant effects, for both Cages 1 and 2 (p < 0.06
and p < 0.02, respectively; Table 9) and the combined cages (p < 0.001, Table 10).
The sign of this effect is dependent upon the parent designated as Pj. The sign of the
effect is a function of the fact that an inverted scale was used and also a reflection of the
relationship between the mid-parent and the means of the F,, F2, and F3 generations
indicating which parent was contributing to the additive variation (51). In Cage 1 (Table
7) the means of the F,, F2, and F3 generations were between the mid-parent (5.73) and
P2. The means of the F2 and F3 in Cage 2 were skewed away from the mid-parent (5.81)
in the same direction, but the F, mean was oriented towards the other parent. Although
the Fj’s mean was on the opposite side, it was within one standard deviation of the mean.
All of the progeny means skewed towards P2, in addition to a significant negative
additive effect indicated that Braxton might be contributing to the additive variation.
Since it has been reported that Braxton does have some level of resistance (64), it stands
to reason that it would contribute to the additive effect.
The full model, Model 2, was then fitted to estimate the epistatic effects as well
as the remaining additive and dominant effects. None of the epistatic (aa, ad, and dd),

60
Table 9. Estimates of the additive, dominant, and epistatic effects in the generation
means for defoliation by soybean looper in the nine populations of the D86-3429 x
Braxton material grown in the field cages at Stoneville, MS in 1991.
Effects
Estimate
Standard
error
t value
Cage 1
Model 1
F2 mean (m)
5.503
0.655
8.399***
Additive (a)
-1.830
0.805
-2.273t
Dominance (d)
-1.287
2.071
-0.621
Model 2
F2 mean (m)
5.839
0.775
7.538**
Additive (a)
0.859
2.525
0.340
Dominance (d)
-2.639
3.103
-0.851
Additive x Additive (aa)
-2.182
2.642
-0.826
Additive x Dominance (ad)
3.212
2.737
1.174
Dominance x Dominance (dd)
3.484
9.370
0.372
Cage 2
Model 1
F2 mean (m)
5.387
0.575
9.374***
Additive (a)
-2.206
0.667
-3.308*
Dominance (d)
-1.385
1.796
-0.771
Model 2
F2 mean (m)
5.613
0.590
9.507**
Additive (a)
-0.031
1.929
-0.016
Dominance (d)
-2.673
2.525
-1.059
Additive x Additive (aa)
-2.191
1.953
-1.122
Additive x Dominance (ad)
2.534
2.063
1.228
Dominance x Dominance (dd)
4.302
7.742
0.556
*,**,“* Significant at the 0.05, 0.01, and 0.001 probability
levels, respectively.
t Significant at the 0.06 probability level.

61
Table 10. Estimates of the additive, dominant, and epistatic effects in the generation
means for defoliation by soybean looper in the nine populations of the D86-3429 x
Braxton material grown in the field cages at Stoneville, MS in 1991.
Effects
Estimate
Standard
error
t value
Caees 1 and 2 Combined
Model 1
F2 mean (m)
5.414
0.503
10.760***
Additive (a)
-2.020
0.487
-4.150***
Dominance (d)
-1.327
1.281
-1.040
Model 2
F2 mean (m)
5.690
0.418
13.630***
Additive (a)
0.424
1.165
0.360
Dominance (d)
-2.669
1.468
-1.820
Additive X Additive {ad)
-2.195
1.197
-1.830
Additive x Dominance (ad)
2.883
1.254
2.300*
Dominance x Dominance (dd)
3.900
4.464
0.870
,*** Significant at the 0.05 and 0.001 probability levels,respectively.

62
additive, or dominant effects were significantly different from zero in either cage.
Although unexpected, ad was significant (p < 0.05) in the combined analysis and neither
the additive or dominant effects were significant. It is interesting and informative that
ad effect is significant. However, it introduces a dimension that cannot be examined
further with currently available data. In Model 1 the epistatic effects are included in the
additive effects. For this reason, Model 1 will be assumed to treat the data adequately.
With an indication that inheritance of resistance to soybean looper was controlled
by an additive genetic system, the data from each cage were tested to fit a two-gene
model. Since no modes were evident in the F2 or F3 populations in either cage (Table
5), Power’s partitioning analysis was employed to analyze the data and test the predicted
gene models following the examples of Deren and Quesenberry (12) and Sage and de
Isturiz (67).
Initially, the frequency distributions of the defoliation data from the parental, Fl5
and F2 populations from each cage were tested for normality (Table 11). As was seen
in 1990, both parental populations had low X2 values which indicated that these
populations were normally distributed about a central mean. In cage 1, the F, population
did not have a normal distribution, but the F2 population’s distribution did fit the normal
curve. The opposite situation occurred with the F, and F2 populations in cage 2. The
differences between the two cages with respect to the F, and F2 populations may be due
to a location effect present within the cages.
In the hypothesized two-gene additive model the middle classes were combined
resulting in three classes for the model to be tested. Partitioning the middle classes

63
Table 11. Goodness of fit test for normality of soybean looper defoliation data from 1991
cage study.
Population
Degrees of Freedom
X2
P
D86-3429
4
Cage 1
1.519
0.90 - 0.75
Braxton
4
7.767
0.25 - 0.10
F,
3
10.667
0.025 - 0.010
f2
8
6.528
0.75 - 0.50
D86-3429
3
Cage 2
0.211
0.990 - 0.975
Braxton
4
4.584
0.500 - 0.250
F,
6
4.000
0.75 - 0.50
f2
8
26.599
0.005>

64
would have been arbitrary, whereas the end class ratios were easily manipulated to test
the model. The F2 data for each cage were tested for goodness of fit to the ratio of
1:14:1 for the two-gene model. The chi-square probabilities were as follows: cage 1,
0.005 > p; and cage 2, 0.10 > p > 0.05. These probabilities are reflective of the test
for normality of the F2 defoliation data in both cages as shown in Table 11. The higher
the probability of a normal distribution of the F2 defoliation data, the lower the
probability that the F2 defoliation data fit the two-gene model. Neither F2 population fit
the two-gene model well. The three-gene model gave a better fit. However, because
there is good reason to believe that the scores assigned in this experiment should not be
assumed to correspond to genotype effects, the Power’s partitioning analysis probably has
no utility in these analyses.
An estimate for the number of genes, n, in this cross contributing to the
expression of resistance to soybean looper was obtained for both cages using a
modification of the formula presented by Poehlman (60). The number of genes
contributing to the expression of resistance to the soybean looper in the cross D86-3429
x Braxton was estimated at n = 1.87 and n = 1.91, cages 1 and 2, respectively.
Broad sense heritability was estimated for resistance to soybean looper in the
progeny from both cages. Although it has already been concluded that resistance to
soybean looper is controlled by additive gene action, narrow-sense heritabilities could not
be calculated because of the size and variances of the BC, and BC2 populations in each
cage. Although broad sense heritabilities were calculated, it is assumed that these
heritabilities are attributed to only the additive genetic variance and are in fact narrow-

65
sense heritabilities. Since the F, generations in both cages exhibited a bimodal
distribution for defoliation scores, the environmental variance was estimated by the
following formula:
(v,/P, + Vp/P2 + vPl<ñ?i )/(vft + vPl + vF)
as was done in estimating the number of genes. This gives a better estimate of the
environmental variance than just using the a2Vx because the number of individuals
comprising the parental generations were relatively large (Table 5). Also, parents are
genetically uniform and should only exhibit variability due to the environment. The
heritabilities were estimated at 62.3% for cage 1 and 64.3% for cage 2. As has been
demonstrated in all previous analyses the results from the two cage studies are in
agreement for the estimate of broad sense heritability. Based on the high heritabilities
for resistance to soybean looper, it is apparent that the phenotypes expressed in this
population were determined by the genetic variation more than by the environmental
variation.
Although the estimates for both cages are in agreement, the estimate is probably
an underestimation of the total number of genes contributing to soybean looper resistance
because these data only estimate the number of genes controlling soybean looper
resistance by which D86-3429 and Braxton differ. In both cages, Braxton exhibited a
low level of resistance with 80% defoliation, whereas Centennial was almost completely
defoliated. When compared to Centennial, which is known for its high susceptibility
(47), Braxton has repeatedly demonstrated 15 to 20% less defoliation. Since the
inception of this study, resistance of Braxton to foliar-feeding insects has been

66
documented by Rowan et al. (64) in a study to compare the resistance of 46
recommended soybean cultivars in Maturity Groups V, VI, VII, and VIII. In that study,
Braxton was cited as being one of the most resistant cultivars in Maturity Group VII.
As in 1990, the insect-resistant germplasm lines, PI 229358 and D75-10169, were
included in both cages to provide additional checks and to document the level of
resistance in the resistant parent, D86-3429. In both cages, D86-3429 was visibly and
statistically (Table 12) more resistant than the original source of resistance, PI 229358.
D86-3429 also exhibited more resistance than did D75-10169 in both cages, although it
was significantly different at p = 0.01 (LSD) only in cage 2 (Table 12). This suggests
that during the five cycles of crossing and selection that occurred in the development of
D86-3429 from PI 229358, additional genes were gained from ’Tracy-M’, a parent in the
fourth cycle. Tracy-M has subsequently been identified as being measurably resistant to
foliar-feeding insects (47).
Summary and Conclusions
This research was conducted to determine the mode of inheritance of resistance
to soybean looper in soybean. The results from Mather’s scaling test (51) applied to the
data from each cage study indicated that generation means depend only on the additive
gene effects for resistance to soybean looper. Utilizing Hayman’s (31) methodology in
terms of the generation-means analysis to analyze the data from each cage separately,
only the additive genetic effects were significant. When the data from the two cages
were combined for analysis the ad epistatic effect was significant (/ = 2.30, p < 0.005).
However, the primary effect is assumed to be additive based on the results of the analysis

67
Table 12. Average defoliation scores in soybean populations resistant to soybean looper in
1991 cage study.
Population
n
Meant
Cage 1
D86-3429
98
3.49 a
D75-10169
28
3.64 a
PI 229358
21
5.13 b
Cage 2
D86-3429
97
3.34 a
D75-10169
27
4.35 b
PI 229358
28
5.42 b
f Means followed by the same letter are not significantly different at the p = 0.01 (LSD).

68
of the data from cage 1 (t = -2.273, p < 0.06), cage 2 (t = -3.308, p < 0.05), and
combined (t = -4.15, p < 0.001), as was indicated by the scaling test. When Power’s
partition analysis (49,61) was applied to the data from each cages the F2 data did not fit
a two-gene additive model for resistance to soybean looper.
A modification of Poehlman’s (60) formula for determination of gene number was
used to estimate the number of genes controlling soybean looper resistance by which the
parents D86-3429 and Braxton differ. Two genes were estimated to contribute to the
expression of resistance to soybean looper in this cross. Since the formula used only
estimates the number of genes in which the two parents, the actual number of genes for
resistance to soybean looper is probably greater than two. It would take further research
with PI 229358 and/or its derived breeding lines to determine exactly how many genes
are controlling resistance to soybean looper. Also, the information generated from this
research could be integrated with the results of research involving other sources of
resistance (36,45,54,72,83) to determine the number of genes that not only control
resistance to soybean looper, but other insect species as well.
Heritability was estimated to be 63% for resistance to soybean looper. Broad
sense heritabilities were calculated because the size and variances of the BC, and BC2
populations were insufficient, thus prohibiting calculation of narrow-sense heritability.
The previous analyses suggests that resistance to soybean looper in this cross is
controlled by two genes which behave additively, therefore it can be assumed that this
heritability is due only to the additive genetic variance. Since the genetic variation

69
influences resistance to soybean looper more than environmental variation, a breeder
should be able to make selections in the F2 population with confidence.
By utilizing a population from a cross between a resistant breeding line and a
highly susceptible breeding line, it should be possible to determine if there are only three
genes for resistance to soybean looper in breeding lines derived from PI 229358. The
single seed descent method of breeding (60) could be used to determine the number of
genes involved by advancing a large population to the F6 generation and evaluating
replicated hills for reaction to soybean looper. For example, assuming three segregating
gene pairs, 91 % of the F6 lines should be homozygous (6) for their respective genotypes,
thus allowing for replications of each genotype. Replicating the F6 lines, as well as their
parents, increases the precision and accuracy of the measured reaction to soybean looper
for each genotype. Examination of results from this type of experiment should indicate
whether or not there are only three genes controlling resistance to soybean looper.
Another way to determine if there are only three genes for resistance to soybean
looper in PI 229358 derived genotypes would be to utilize restriction fragment length
polymorphism (RFLP) markers. RFLP markers have been widely used in other plant
species to construct linkage maps and to place genes that control qualitative and
quantitative traits onto these maps. Recently, RFLP markers have been used to map
phytophthora resistance loci in soybean (13). This methodology could be used to map
the genes and possibly determine any linkage groups that may be associated with the
genes controlling resistance to soybean looper.

70
Since the results of this study have shown that resistance to soybean looper was
a quantitative trait that was additive, it should be possible to increase the level of
resistance in future breeding lines. This might be accomplished by utilizing new sources
of resistance which have additional genes for insect resistance. Another method of
increasing the level of resistance to soybean looper would be by interspecific
hybridization which would allow genes for resistance in other species to be incorporated
into the soybean plant.
To fully exploit the benefits of insect resistant breeding lines, it will become
necessary to understand the mechanisms of resistance that the plants possess. Depending
on the source of resistance, the mechanism of resistance may be different in the various
sources. Several researchers (39,53,63,73,81) have already conducted studies on the
mechanisms of resistance, but further research is needed. As new technology becomes
available, the mysteries of the mechanisms of insect resistance and the genes that control
their expression are closer to being solved.

CHAPTER 5
SUMMARY AND CONCLUSIONS
Soybean is grown worldwide primarily for its oil and meal products, and to a
lesser degree the whole bean product. As in any crop, soybean is exposed to biotic and
abiotic yield-reducing factors such as diseases, nematodes, insects, and drought
throughout the growing season. To produce a profitable crop a grower must modify the
effects of these yield-reducing factors by one or more of the methods described by
Metcalf and Luckmann (50). Environmentally, the best approach to any pest problem
would be to utilize genes for resistance in available germplasm to develop resistant
cultivars (82).
Soybean looper is a major pest to soybean in many areas of the southeastern
United States. The soybean looper, more so than any other arthropod pest, has become
increasingly resistant to available insecticides (10,18,48,58,60) making it difficult to
control. By utilizing soybean germplasm lines that possess genes for resistance to foliar-
feeding insects, it is possible to increase the host plant responses for control as has been
demonstrated by the development of the insect-resistant cultivars Lamar (28) and
Crockett (5).
The objectives of this study were: (i) to evaluate three methods of rating for
defoliation of soybean by soybean looper, and (//) to determine the inheritance of
71

72
resistance in soybean to the soybean looper. The latter was the most important objective
since knowledge of inheritance of a particular trait increases the efficiency for
incorporating genes for that trait into an agronomically desirable plant type. Rating
methodology can affect the accuracy of classification and therefore was evaluated
separately to offer guidelines for future insect defoliation studies.
Evaluation of Rating Methods
The three methods of defoliation assessment evaluated in 1990 and 1991 were the
whole plant visual rating (WP), the partitioned plant average visual rating (AV), and
measured average leaf area of the partitioned plant (AL). The WP and AV were visual
assessments of defoliation based on a l-to-10 scale divided into 10% increments with a
score of 1 signifying defoliation of 0 to 10% and a score of 10 signifying defoliation of
91 to 100%. The WP method was based on one observation per plant, whereas the AV
method was based on the mean of three observations per plants. The absolute method
(AL) utilized a LI-COR LI-3000 leaf area meter to measure the area of three trifoliolates,
one each from the top, middle, and bottom of the plant.
In 1990 no absolute method comparisons were possible, however the WP and AV
were highly correlated (r=0.65, p < 0.001). Adjustments in the method of calculating
percent defoliation by AL for 1991 allowed for direct comparisons with the WP and AV
methods. All correlations (WP vs. AV, WP vs. AL, and AV vs. AL) were significant
at the p < 0.001 level whether the cages were examined separately or pooled. The
results of these correlations indicate that the evaluator could obtain similar estimates of
defoliation regardless of which of the three methods is utilized. These findings are

73
similar to the results of another relative versus absolute estimation of defoliation study
conducted in 1985 by G.R. Bowers (G.R. Bowers, 1992, personal communication).
A paired difference comparison was performed to test the equality of the
defoliation assessment between the relative methods in 1990 and among the three
methods in 1991. The results indicated that estimates of defoliation by the two relative
methods in 1990 and by all three in 1991 were significantly different. With the relative
methods, WP and AV, the evaluator estimated somewhat similar amounts of defoliation
across all plants when compared with each other, even though the difference between the
two methods was highly significant. However, in comparisons with an absolute method,
as in 1991, it becomes apparent that the evaluator’s estimates of defoliation with the WP
method were closer to the actual than were the estimates of the AV method. These
findings were validated with McNemar’s test for concordance.
The relative variation (RV) was calculated for each method as a measure of
precision. In 1990 the precision of the WP and AV methods were essentially equal.
Similar results were obtained in 1991 by all three methods. Both years the precision of
the methods was essentially equal and well below 10, the value that Pedigo states as
being desirable in research. These results indicate that either relative method, WP or
AV, would be suitable for estimating defoliation by insects.
Based on the results of these studies it is apparent that relative estimates of
defoliation can be used in genetic studies or routine screenings of advanced breeding
lines, in lieu of an absolute estimate. Since the two relative methods yielded similar
results it would advantageous to use the simplest and most economical method for

74
making estimates of defoliation. With increased experience in visual estimates of
defoliation, researchers should be able to estimate defoliation with reliability and
repeatability.
Inheritance Study
This study was undertaken to determine the mode of inheritance of resistance to
the soybean looper in soybean. The study consisted of two years of field cage
evaluations of progeny derived from the cross D86-3429 x Braxton. The results of the
preliminary study in 1990 suggested that the genes controlling resistance to soybean
looper might be inherited quantitatively. The 1991 experiment was designed so that
inheritance of resistance to soybean looper could be examined quantitatively as well as
qualitatively. Also, the 1991 experiment was duplicated in two field cages so that the
experiment would be repeated in two separate environments to strengthen the results and
conclusions of the study.
In the 1991 study, estimates of defoliation were made for each plant using the AV
method. Although this method was more time consuming than the WP method, it was
chosen because its estimate of defoliation was based on three observations per plant
rather than one observation. After assessing each plant in each cage for defoliation by
the soybean looper the data generated were analyzed by several methods of genetic
analysis.
Mather’s scaling test (51) was applied to the defoliation data from each cage study
and it was determined that generation means depend only on the additive gene effects and
no epistasis was present. Utilizing Hayman’s (31) methodology in terms of the

75
generation-means analysis to analyze the data from each cage separately, only the
additive genetic effects were significant. When the data from the two cages were
combined for analysis the ad epistatic effect was significant {t = 2.30, p < 0.005).
However, the primary effect is assumed to be additive based on the results of the analysis
of the data from cage 1 (t = -2.273, p < 0.06), cage 2 (t = -3.308, p < 0.05), and
combined {t = -4.15, p < 0.001), as was indicated by the scaling test. Applying
Power’s partitioning analysis (49,61) to the data from each cage it was determined that
in this cross the F2 data did not fit the two-gene additive model for resistance to soybean
looper. There is good reason to believe that the defoliation scores assigned in this
experiment should not be assumed to correspond to genotype effects, therefore the
Power’s partitioning analysis has no utility in these analyses.
Based on a modification of the formula presented by Poehlman (60), it was
determined that two genes contribute to the expression of resistance to soybean looper
in this cross. However this formula only estimates the number of genes for resistance
to soybean looper by which the two parents differ. In actuality the number of genes for
resistance to soybean looper is probably greater than two, but it would take further
research to determine this.
Broad sense heritability based on the defoliation data was estimated to be 63 % for
resistance to soybean looper. Since the previous analyses suggest that resistance to the
soybean looper is controlled by two genes which behave additively, it can be assumed
that this heritability is due only to the additive genetic variation. A heritability of this

76
magnitude suggests that a breeder should be able to make progress by selecting for
resistance to soybean looper in early generations.
This research could be continued with breeding lines derived from PI 229358 or
by incorporating additional sources of resistance, such as PI 227687, PI 171451, PI
417061, or PI 507301 into agronomically adapted breeding lines. These elite genotypes
could then be cross-pollinated with a highly susceptible cultivar, such as Centennial, to
develop large F2 populations for future evaluations. In order to determine the number
of genes controlling resistance to the soybean looper and/or other species, these F2
populations could be screened by conventional methods or some of the newer methods,
such as RFLP analysis.
Since resistance to soybean looper in this study was determined to be additive, it
should be possible to increase resistance in future cultivars by pyramiding additional
genes from other sources of resistance. This can be done by conventional plant breeding
techniques by utilizing soybean germplasm lines that show resistance to leaf-feeding
insects and incorporating these genes into adapted cultivars or advanced breeding lines.
The same goals could also be accomplished by any of the new methods of biotechnology
that are currently available.
Additional research will need to be conducted on sources of resistance other than
PI 229358-derived lines to determine how many other genes for resistance are available
to breeders for utilization in their breeding programs. In order to fully exploit the
benefits of insect resistance in a breeding program, it will become necessary to
understand the mechanisms of resistance as well as the underlying genes that code for the

77
mechanism. As new technology becomes available, the mysteries of the mechanisms of
insect resistance will be solved. Ideally, the future promises the development of a
resistant cultivar that has many genes for resistance that code for all three mechanisms
of resistance.

APPENDIX A
LIST OF DEFOLIATION SCORES GENERATED BY WHOLE PLANT VISUAL,
PARTITIONED PLANT VISUAL, AND LEAF AREA OF THE PARTITIONED
PLANT RATING METHODS IN 1990.

79
List of defoliation scores generated by whole plant visualf, partitioned plant visualf, and leaf
area of the partitioned plants rating methods.
Whole Avg. Avg.
Observation Plant Plant Visual Leaf Area
1
BBE-5-3
1
2.67
57.55
2
BBE-5-4
2
2.00
66.45
3
BBE-5-5
4
1.67
72.74
4
BBE-5-6
3
2.33
65.45
5
BBE-5-7
3
3.00
81.61
6
BBE-5-8
2
2.33
75.80
7
BBE-5-9
2
2.33
111.43
8
BBE-5-10
3
2.67
66.73
9
BBE-6-1
4
3.00
84.04
10
BBE-6-2
2
2.33
47.86
11
BBE-6-3
2
2.33
66.22
12
BBE-6-4
4
2.33
32.03
13
BBE-6-5
3
2.00
87.55
14
BBE-6-6
3
2.67
86.85
15
BBE-6-7
2
2.33
52.67
16
BBE-6-8
4
3.33
76.31
17
BBE-6-9
4
3.33
71.35
18
BBE-6-10
4
2.00
64.61
19
BBE-7-1
3
4.33
100.27
20
BBE-7-2
3
3.00
71.86
21
BBE-7-3
4
3.33
68.00
22
BBE-7-4
3
2.67
75.74
23
BBE-7-5
4
1.33
61.07
24
BBE-7-6
3
2.00
77.64
25
BBE-7-7
4
2.67
67.08
26
BBE-7-8
4
4.33
74.19
27
BBE-7-9
2
1.67
73.05
28
BBE-7-10
4
2.00
63.85
29
BBE-8-1
3
3.33
58.87
30
BBE-8-2
3
1.00
88.32
31
BBE-8-3
2
2.00
72.49
32
BBE-8-4
4
2.00
70.66
33
BBE-8-5
4
2.67
73.00
34
BBE-8-6
4
2.00
79.00

35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
80
Plant
Whole
Plant
Avg.
Visual
Avg.
Leaf Area
BBE-8-7
3
2.00
80.88
BBE-8-8
3
1.67
94.90
BBE-8-9
2
2.00
60.38
BBE-8-10
4
2.67
31.06
BBE-9-1
4
4.67
70.52
BBE-9-2
3
3.67
66.67
BBE-9-3
5
5.00
47.64
BBE-9-4
4
5.00
69.57
BBE-9-5
4
5.33
57.64
BBE-9-6
3
3.00
87.20
BBE-9-7
4
2.67
74.37
BBE-9-8
3
1.67
75.70
BBE-9-9
3
2.67
66.17
BBE-9-10
4
3.67
74.00
BBE-10-1
2
1.33
45.08
BBE-10-2
2
2.67
77.06
BBE-10-3
2
3.67
63.51
BBE-10-4
2
3.00
67.74
BBE-10-5
2
1.67
71.65
BBE-10-6
2
2.67
70.88
BBE-10-7
2
2.00
76.32
BBE-10-8
3
2.67
89.68
BBE-10-9
3
1.33
52.45
BBE-10-10
3
3.33
29.95
BBE-11-1
3
7.00
64.62
BBE-11-2
3
4.67
75.10
BBE-11-3
3
4.00
55.21
BBE-11-4
3
3.67
65.95
BBE-11-5
3
4.33
62.94
BBE-11-6
2
3.33
88.50
BBE-11-7
3
3.67
74.01
BBE-11-8
3
5.33
63.24
BBE-11-9
2
1.67
97.20
BBE-11-10
4
6.67
74.45
BBE-12-1
3
4.33
62.75
BBE-12-2
3
2.33
78.93

71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
81
Plant
Whole
Plant
Avg.
Visual
Avg.
Leaf Area
BBE-12-3
3
2.67
72.66
BBE-12-4
3
2.00
92.53
BBE-12-5
3
2.33
93.98
BBE-12-6
2
2.33
106.10
BBE-12-7
3
4.33
72.10
BBE-12-8
3
3.00
86.99
BBE-12-9
3
4.33
98.32
BBE-12-10
2
4.00
60.94
BBE-13-1
4
5.00
53.08
BBE-13-2
2
1.67
91.42
BBE-13-3
2
1.67
73.57
BBE-13-4
4
4.00
75.45
BBE-13-5
3
5.00
46.18
BBE-13-6
4
4.00
63.52
BBE-13-7
3
4.00
68.61
BBE-13-8
3
2.33
91.37
BBE-13-9
3
3.00
51.28
BBE-13-10
4
5.00
58.01
BBE-14-1
3
4.67
57.90
BBE-14-2
3
5.00
65.62
BBE-14-3
2
4.33
65.58
BBE-14-4
2
2.00
83.45
BBE-14-5
4
5.00
57.62
BBE-14-6
3
3.67
84.46
BBE-14-7
4
6.33
58.82
BBE-14-8
4
4.67
50.44
BBE-14-9
2
3.00
79.47
BBE-14-10
5
5.00
42.37
BBE-4-1
3
2.67
69.64
BBE-4-2
3
2.67
72.35
BBE-4-3
4
3.33
74.24
BBE-4-4
3
2.00
80.57
BBE-4-5
3
4.67
49.60
BBE-4-6
5
5.67
60.68
BBE-4-7
4
4.33
80.55
BBE-4-8
2
2.33
72.02

82
Observation
Plant
Whole
Plant
Avg.
Visual
Avg.
Leaf Area
107
BBE-4-9
3
2.67
58.72
108
BBE-4-10
3
5.67
65.01
109
BBE-1-1
5
6.67
77.55
110
BBE-1-2
3
2.33
92.04
111
BBE-1-3
3
2.67
72.01
112
BBE-1-4
4
5.00
63.01
113
BBE-1-5
4
5.67
64.55
114
BBE-1-6
4
6.00
51.66
115
BBE-1-7
4
5.33
82.70
116
BBE-1-8
3
2.67
64.59
117
BBE-1-9
3
4.33
60.43
118
BBE-1-10
4
3.67
56.59
119
BBE-3-1
2
3.33
69.57
120
BBE-3-2
2
3.33
64.69
121
BBE-3-3
3
2.67
69.09
122
BBE-3-4
2
1.67
60.31
123
BBE-3-5
2
3.00
82.66
124
BBE-3-6
2
2.33
68.51
125
BBE-3-7
2
2.33
70.63
126
BBE-3-8
2
2.33
65.65
127
BBE-3-9
3
2.67
80.47
128
BBE-3-10
3
2.33
69.33
129
BBE-2-1
2
2.00
80.21
130
BBE-2-2
2
2.33
81.01
131
BBE-2-3
2
2.00
67.84
132
BBE-2-4
2
2.67
52.42
133
BBE-2-5
2
2.00
73.62
134
BBE-2-6
3
4.00
82.98
135
BBE-2-7
3
2.33
77.75
136
BBE-2-8
2
2.00
69.91
137
BBE-2-9
2
2.00
64.00
138
BBE-2-10
3
3.00
79.92
139
BBE-16-1
4
5.33
51.63
140
BBE-16-2
4
6.33
45.94
141
BBE-16-3
5
6.00
40.06
142
BBE-16-4
4
4.00
62.02

83
Observation
Plant
Whole
Plant
Avg.
Visual
Avg.
Leaf Area
143
BBE-16-5
5
5.67
45.71
144
BBE-16-6
6
7.00
35.54
145
BBE-16-7
5
4.67
75.60
146
BBE-16-8
7
5.67
37.47
147
BBE-16-9
4
5.33
59.21
148
BBE-16-10
7
7.33
62.08
149
BBE-15-1
5
5.33
74.96
150
BBE-15-2
3
4.00
62.52
151
BBE-15-3
4
4.67
64.42
152
BBE-15-4
3
3.67
67.52
153
BBE-15-5
3
3.33
76.06
154
BBE-15-6
4
5.33
50.96
155
BBE-15-7
4
5.00
70.10
156
BBE-15-8
3
4.33
74.26
157
BBE-15-9
4
4.67
64.68
158
BBE-15-10
4
5.67
46.88
159
BBE-17-1
4
6.00
44.94
160
BBE-17-2
5
6.00
39.03
161
BBE-17-3
3
4.00
80.14
162
BBE-17-4
4
6.00
39.94
163
BBE-17-5
3
4.33
52.10
164
BBE-17-6
4
5.33
42.98
165
BBE-17-7
3
5.33
78.97
166
BBE-17-8
4
4.67
68.50
167
BBE-17-9
4
4.33
52.05
168
BBE-17-10
4
5.33
58.14
169
BBE-18-1
4
5.67
56.14
170
BBE-18-2
4
5.33
61.43
171
BBE-18-3
5
5.00
51.84
172
BBE-18-4
4
5.33
54.93
173
BBE-18-5
4
5.33
58.81
174
BBE-18-6
5
5.33
53.30
175
BBE-18-7
4
5.00
75.94
176
BBE-18-8
4
3.00
62.40
177
BBE-18-9
4
5.33
61.60
178
BBE-18-10
4
5.00
78.89

84
Observation
Plant
Whole
Plant
Avg.
Visual
Avg.
Leaf Area
179
BBE-19-1
5
5.67
56.32
180
BBE-19-2
3
3.33
74.48
181
BBE-19-3
5
4.00
49.36
182
BBE-19-4
5
6.00
58.93
183
BBE-19-5
3
4.67
55.07
184
BBE-19-6
3
4.33
88.80
185
BBE-19-7
5
5.67
50.52
186
BBE-19-8
4
5.67
57.10
187
BBE-19-9
4
6.33
48.33
188
BBE-19-10
4
3.00
74.01
189
BBE-20-1
4
3.67
53.10
190
BBE-20-2
3
2.67
71.80
191
BBE-20-3
4
4.33
63.00
192
BBE-20-4
3
4.00
60.55
193
BBE-20-5
2
2.67
66.57
194
BBE-20-6
4
3.33
69.70
195
BBE-20-7
3
3.33
72.31
196
BBE-20-8
4
3.67
56.56
197
BBE-20-9
3
4.33
64.67
198
BBE-20-10
4
2.67
53.80
Mean (n= 198)
3.32
3.68
67.13
Variance
0.98
2.07
201.92
Std.Dev.
0.99
1.44
14.21
t Based on a 1 to 10 scale divided into 10% increments with a 10 reflecting 91-100% defoliation.
$ LI-COR LI-3000 portable leaf area meter measurements in cm2 of leaf tissue.

APPENDIX B
LIST OF DEFOLIATION SCORES GENERATED BY WHOLE PLANT VISUAL,
PARTITIONED PLANT VISUAL, AND LEAF AREA OF THE PARTITIONED
PLANT RATING METHODS IN 1991.

86
List of defoliation scores! generated by whole plant visual, partitioned plant visual, and leaf area
of the partitioned plant! rating methods.
Observation
Cage
Plant
Whole plant
Avg.
visual
Avg.
LICOR
1
1
9
3
2.00
1.67
2
1
197
5
4.00
2.00
3
1
214
4
3.00
3.00
4
1
15
8
7.00
5.00
5
1
250
-§
-
-
6
1
235
4
4.33
2.00
7
1
63
6
6.67
5.00
8
1
180
2
2.00
1.33
9
1
61
3
3.33
2.67
10
1
42
4
3.67
2.00
11
1
223
3
3.67
4.00
12
1
150
6
5.00
4.00
13
1
186
8
7.67
5.00
14
1
26
2
3.00
2.33
15
1
225
3
2.67
2.33
16
1
187
7
5.67
4.67
17
1
3
2
3.33
1.33
18
1
248
8
6.67
6.33
19
1
17
9
7.67
7.00
20
1
196
8
8.00
6.00
21
1
222
4
3.67
3.00
22
1
275
5
7.00
5.00
23
1
8
2
3.67
1.67
24
1
158
7
5.67
4.33
25
1
166
7
7.67
4.00
26
1
140
8
8.00
5.67
27
1
141
4
5.33
2.33
28
1
138
2
3.67
2.33
29
1
85
5
5.00
3.33
30
1
244
4
4.33
3.00
31
1
122
3
5.33
2.00
32
1
144
8
8.33
6.00
33
1
105
5
5.33
3.00
34
1
92
6
6.67
4.67
35
1
126
8
8.33
6.00
36
1
67
7
7.00
5.00
37
1
51
6
8.67
6.00
38
1
75
7
8.33
4.67
39
1
80
5
6.33
3.33

87
Observation
Cage
Plant
Whole plant
Avg.
visual
Avg.
LICOR
40
1
178
4
5.33
3.33
41
1
179
5
6.00
4.00
42
1
22
8
7.67
4.67
43
1
110
7
6.33
5.00
44
1
270
9
8.67
7.00
45
1
212
-
-
-
46
1
100
6
5.67
2.67
47
1
213
3
4.33
2.00
48
1
25
2
4.00
3.00
49
1
103
5
5.67
2.67
50
1
79
3
3.67
3.00
51
2
248
4
3.33
1.67
52
2
213
2
3.33
3.33
53
2
17
8
8.33
5.67
54
2
270
8
9.00
4.33
55
2
150
5
5.67
4.00
56
2
100
4
5.00
3.00
57
2
194
6
6.67
3.33
58
2
186
6
5.67
4.00
59
2
61
7
7.67
5.00
60
2
79
4
4.00
2.67
61
2
9
4
3.00
4.33
62
2
250
9
9.33
6.33
63
2
26
2
2.33
1.67
64
2
222
7
6.00
5.00
65
2
214
1
3.33
2.00
66
2
244
2
4.00
2.00
67
2
103
5
6.33
3.00
68
2
178
3
4.67
3.00
69
2
22
8
8.67
5.33
70
2
85
7
5.33
5.33
71
2
180
3
5.00
2.67
72
2
42
3
6.33
5.00
73
2
110
5
4.67
3.33
74
2
212
-
-
-
75
2
275
5
5.00
3.33
76
2
158
6
7.00
3.00
77
2
51
4
4.33
3.00
78
2
75
5
6.33
3.33
79
2
25
5
2.00
1.67
80
2
8
3
3.00
1.67

88
Observation
Cage
Plant
Whole plant
Avg.
visual
Avg.
LICOR
81
2
140
8
4.67
2.33
82
2
15
3
8.67
5.00
83
2
138
3
2.67
1.33
84
2
67
6
6.67
3.33
85
2
126
7
8.00
6.00
86
2
63
8
7.67
5.33
87
2
179
6
7.33
5.33
88
2
80
7
7.67
2.67
89
2
92
6
6.67
2.00
90
2
3
3
4.33
2.00
91
2
223
4
6.67
3.33
92
2
166
5
7.00
4.00
93
2
122
2
3.67
2.33
94
2
225
3
4.67
2.00
95
2
196
9
9.00
5.33
96
2
144
9
9.00
5.00
97
2
187
3
5.33
2.33
98
2
105
4
6.33
3.00
99
2
141
3
4.33
3.00
100
2
235
5
7.67
3.00
Mean (N = 97)
5.10
5.65
3.62
Variance
4.70
3.88
2.19
Std.Dev.
2.17
1.97
1.48
t Based on a 1 to 10 scale divided into 10% increments with a 10 reflecting 91-100% defoliation.
$ Based on LI-COR LI-3000 portable leaf area meter measurements that have been converted to %
defoliation and categorized into the l-to-10 scale.
§ Missing data points.

APPENDIX C
GENERATION MEANS ANALYSIS
SAS PROGRAM
OPTIONS LS=80 PS=60;
/*This is a Generation Means Analysis based on Hayman’s methodology, Heredity
12:371-390 */
DATA;
INFILE ’ArCagel.dat’;
/* The INPUT statement will vary according to the data set, you need generation
("GEN") and a dependent variable */
INPUT row plant gen $ fc$ rating 1 rating2 rating3;
MEANRATE = (rating 1+rating2+rating3)/3;
PROC SORT; BY GEN;
PROC MEANS; BY GEN; VAR MEANRATE;
OUTPUT OUT=NE MEAN=Y VAR=V STDERR=S;
DATA NEW; SET NE; RS = 1/S; IF GEN=’D’ OR GEN=’BD’ THEN DELETE;
PROC PRINT;
/* You need a minimum of 6 generations to conduct this analysis, if the number of
generations used are different than the amount (9) in this example, refer to Gamble’s
paper (Canadian J. Plant Sci. 42:339-348) to obtain the proper coefficients. Another
source of information is Jennings, et al. (IA State J. Research, 48:267-280) */
DATA COEFCNTS;
INPUT GEN $ XI X2 X3 X4 X5;
CARDS;
BC1 0.5 0.0 0.25 0.0 0.0
BC2 -0.5 0.0 0.25 0.0 0.0
BS1 0.5 -0.25 0.25 -0.25 0.0625
BS2 -0.5 -0.25 0.25 0.25 0.0625
FI 0.0 0.5 0.0 0.0 0.25
F2 0.0 0.0 0.0 0.0 0.0
F3 0.0 -0.25 0.0 0.0 0.0625
PI 1.0-0.5 1.0 -1.0 0.25
P2 -1.0-0.5 1.0 1.0 0.25
DATA FINAL; MERGE NEW COEFCNTS;
PROC PRINT;
89

90
/* This model tests the significance of the additive (XI) and the dominant (X2) effects
*/
TITLE ’PROC REG WEIGHTED’;
PROC REG;
MODEL Y = XI X2;
WEIGHT RS;
/* The model must be weighted to account for the unequal population sizes among the
generations. See Rowe and Alexander, Crop Sci. 20:109-110, for further detail. The next
model tests for the epistatic effects, additive-additive (X3), additive-dominant (X4), and
dominant-dominant (X5), as well as the additive (XI) and dominant (X2) effects. */
PRO REG;
MODEL Y = XI X2 X3 X4 X5;
WEIGHT RS;
RUN;
/* This next set of models are tested with PROC GLM instead of PROC REG */
TITLE ’PROC GLM WEIGHTED’;
PROC GLM;
MODEL Y = XI X2;
WEIGHT RS;
PROC GLM;
MODEL Y = XI X2 X3 X4 X5;
WEIGHT RS;
RUN;
/* The next series of models are the same as the previous models, except that they
assume equal population size among the generations, and are therefore not weighted. */
TITLE ’PROC REG NOWEIGHT’;
PROC REG;
MODEL Y = XI X2;
PROC REG;
MODEL Y = XI X2 X3 X4 X5;
RUN;
TITLE ’PROC GLM NOWEIGHT’;
MODEL Y = XI X2;
PROC GLM;
MODEL Y = XI X2 X3 X4 X5;
RUN;
/* Depending on whether or not the population sizes among the generations are equal,
report the model that has the most significant effects from either PROC GLM or PROC
REG. */

APPENDIX D
GENERATION MEANS ANALYSIS
SAS PROGRAM
/* This is a Generation Means Analysis that is based on Hayman’s methodology,
Heredity 12:371-390. This analysis tests for block effects as well as estimating the
additive, dominant, and epistatic effects. */
DATA A;
INFILE ’A:CAGE1.DAT’;CAGE= 1;
/* The INPUT statement will vary according to the data set, you need generation
("GEN") and a dependent variable. */
INPUT ROW PLANT GEN $ FC $ RATING 1 RATING2 RATING3;
MEANRATE = (RATING 1+RATING2+RATING3)/3;
DATA B;
INFILE ’A:CAGE2.DAT’;CAGE=2;
INPUT ROW PLANT GEN $ FC $ RATING 1 RATING2 RATING3;
MEANRATE = (RATING 1+RATING2 + RATING3)/3;
DATA BOTH;SET A B;RUN;
PROC SORT; BY CAGE GEN;
PROC MEANS; BY CAGE GEN; VAR MEANRATE;
OUTPUT OUT=NE MEAN = Y VAR = V STDERR = S;
DATA NEW; SET NE; RS = l/S;IFGEN=’D’OR GEN=’BD’ THEN DELETE; PROC
PRINT;
/* You need a minimum of 6 generations to conduct this analysis, if the number of
generations used are different than the amount (9) in this example, refer to Gamble’s
paper (Canadian J. Plant Sci. 42:339-348) to obtain the proper coefficients. Another
source of information is Jennings, et al. (IA State J. Research, 48:267-280) */
DATA COEFCNTS;
INPUT CAGE GEN $ XI X2 X3 X4 X5;
CARDS;
1 BC1 0.5 0.0 0.25 0.0 0.0
1 BC2 -0.5 0.0 0.25 0.0 0.0
1 BS1 0.5 -0.25 0.25 -0.25 0.0625
1 BS2 -0.5 -0.25 0.25 0.25 0.0625
1 FI 0.0 0.5 0.0 0.0 0.25
1 F2 0.0 0.0 0.0 0.0 0.0
1 F3 0.0-0.25 0.0 0.0 0.0625
91

92
1 PI 1.0-0.5 1.0 -1.0 0.25
1 P2 -1.0-0.5 1.0 1.0 0.25
2 BC1 0.5 0.0 0.25 0.0 0.0
2 BC2 -0.5 0.0 0.25 0.0 0.0
2 BS1 0.5 -0.25 0.25 -0.25 0.0625
2 BS2 -0.5 -0.25 0.25 0.25 0.0625
2 FI 0.0 0.5 0.0 0.0 0.25
2 F2 0.0 0.0 0.0 0.0 0.0
2 F3 0.0 -0.25 0.0 0.0 0.0625
2 PI 1.0-0.5 1.0 -1.0 0.25
2 P2 -1.0-0.5 1.0 1.0 0.25
PROC SORT DATA=COEFCNTS;BY CAGE GEN;
DATA FINAL; MERGE NEW COEFCNTS;BY CAGE GEN;
PROC PRINT;
/* This model tests the significance of the additive (XI) and dominant (X2) effects */
PROC GLM DATA=FINAL;BY CAGE;
MODEL Y = XI X2 /SSI SS3;
WEIGHT RS;
/* The model must be weighted to account for the unequal population sizes among the
generations. See Rowe and Alexander, Crop Sci. 20:109-110, for further detail. */
PROC GLM DATA=FINAL;CLASS CAGE;
MODEL Y = CAGE XI X2 CAGE*X1 CAGE*X2/SS1 SS3 SOLUTION;
WEIGHT RS;
PROC GLM DATA=FINAL;CLASS CAGE;
MODEL Y = CAGE XI X2 /SSI SS3 SOLUTION;
WEIGHT RS;
/* The next model tests for the epistatic effects, additive-additive (X3), additive-dominant
(X4), and dominant-dominant (X5), as well as the additive (XI) and dominant (X2)
effects. */
PROC GLM;BY CAGE;
MODEL Y = XI X2 X3 X4 X5;
WEIGHT RS;
RUN;
PROC GLM;CLASS CAGE;
MODEL Y = CAGE XI X2 X3 X4 X5 CAGE*X1 CAGE*X2 CAGE*X3 CAGE*X4
CAGE*X5/SS1 SS3 SOLUTION;
WEIGHT RS;
RUN;
PROC GLM;CLASS CAGE;
MODEL y = CAGE XI X2 X3 X4 X5 /SSI SS3 SOLUTION;
WEIGHT RS;
RUN;

APPENDIX E
TEST FOR NORMALITY
x = [1(0.5) + 33(1.5) + 78(2.5) + 154(3.5) + 39(4.5) + 6(5.5) +
2(6.5) + 1(7.5)]/314
= 3.2
s2 = [l(0.5-3.2)2 + 33(1.5-3.2)2 + 78(2.5-3.2)2 + 159(3.5-3.2)2 +
39(4.5-3.2)2 + 6(5.5-3.2)2 + 2(6.5-3.2)2 + l(7.5-3.2)2]/(314-l)
= 0.9343
s = (0.9343)*
= 0.96662
Next, the value x was calculated for each class as follows:
Class 1
x = (3.2
Class 2
x = (3.2
Class 3
x = (3.2
Class 4
x = (3.2
Class 5
x = (3.2
Class 6
x = (3.2
Class 7
x = (3.2
Class 8
x = (3.2
0.5)/0.96662 = 2.30701
1.5)/0.96662 = 1.27248
2.5)/0.96662 = 0.23794
3.5)/0.96662 = -0.79659
4.5)/0.96662 = -1.83112
5.5)/0.96662 = -2.86566
6.5)/0.96662 = -3.90019
7.5)/0.96662 = -4.93472
The corresponding Z value for each class was obtained by finding x in a table of
probabilities of the normal distribution (76). The Z value obtained for each class was
then subtracted from one and multiplied by 100% to obtain the theoretical % of the
population that fell into each class. The theoretical number of each class was obtained
by multiplying the total n of the population by the theoretical % of each class. This was
done for each class as shown below:
93

94
X
Z
1.000 - Z
Theoretical
% Class
Theoretical
Distribution
2.30701
0.99111
0.00889
0.89
3
1.27248
0.89796
0.10204
9.32
29
0.23794
0.59484
0.40516
30.31
95
-0.79659
0.21185
0.78815
38.30
120
-1.83112
0.03362
0.96638
17.82
56
-2.86566
0.00205
0.99795
3.16
10
-3.90019
0.00005
0.99995
0.2
1
-4.93472
0.00001
0.99999
0.004
0
The observed versus theoretical distribution of the F2 population was then tested with a
X2 test.

REFERENCES
1. Allard, R.W. 1960. Principles of plant breeding. John Wiley and Sons, Inc. New
York, NY.
2. Anonymous. 1979. USDA agricultural statistics. U.S. Gov. Print. Office,
Washington, D.C.
3. Anonymous. 1990. USDA agricultural statistics. U.S. Gov. Print. Office,
Washington, D.C.
4. Berger, R.D. 1988. Measuring disease intensity, p.1-4. In Biological and cultural
tests for control of plant diseases. American Phytopathological Society Press, St.
Paul, MN.
5. Bowers, G.R. 1990. Registration of ’Crockett’ soybean. Crop Sci. 30:427.
6. Briggs, F.N., and P.F. Knowles. 1967. Introduction to plant breeding. Reinhold
Publishing Corporation, .
7. Campbell, C.L., and L.V. Madden. 1990. Introduction to plant disease
epidemiology. John Wiley and Sons, New York, NY.
8. Chapman, S.R., and F.H. McNeal. 1970. Gene effects for protein in five spring
wheat crosses. Crop Sci. 10:45-46.
9. Chiang, H.S., D.M. Norris, A. Ciepiela, A. Oosterwyk, P. Shapiro, and M.
Jackson. 1986. Comparitive constitutive resistance in soybean lines to Mexican
bean beetle. Entomol. Exp. Appl. 42:19-26.
10. Chiu, P. S-B., and M.H. Bass. 1978. Soybean looper: susceptibility of larvae to
insecticides. J. Ga. Entomol. Soc. 13:169-173.
11. Cooper, R.L., and R.B. Hammond. 1988. Registration of Mexican bean beetle
resistant germplasm line HC83-1239. Crop Sci. 28:1037-1038.
95

96
12. Deren, C.W., and K.H. Quesenberry. 1989. Inheritance of photoperiod-induced
flowering in three photoperiodic lines of Aeschynomene americana L. Theor.
Appl. Genet. 78:825-830.
13. Diers, B.W., L. Mansur, J. Imsande, and R.C. Shoemaker. 1992. Mapping
phytophthora resistance loci in soybean with restriction fragment length
polymorphism markers. Crop Sci. 32:377-383.
14. Ennis, W.B., Jr. 1976. Modem methods for controlling pests, p. 375-386. In
L.D. Hill (ed.) World soybean research. Interstate Printers and Publishers.
Danville, IL.
15. Entomological Society of America, Southeastern Branch, Insect Detection,
Evaluation, and Predication Committee. 1984. Insect detection, evaluation and
prediction report [for] 1982, v.7. MS State, MS.
16. Entomological Society of America, Southeastern Branch, Insect Detection,
Evaluation, and Predication Committee. 1985. Insect detection, evaluation and
prediction report [for] 1984, v.9. Clemson, SC.
17. Fehr, W.R., and C.E. Caviness. 1977. Stages of soybean development. Agrie,
and Home Economics Exp. Stn. and Cooperative Ext. Serv., Iowa State Univ.
and Arkansas Agrie. Exp. Stn. Spec. Rep. 80.
18. Felland, C.M., H.N. Pitre, R.G. Luttrell, and J.L. Hamer. 1990. Resistance to
pyrethroid insecticides in soybean looper (Lepidoptera:Noctuidae) in Mississippi.
J. Econ. Entomol. 83:35-40.
19. Funderburk, J.E., A.R. Soffes, R.D. Barnett, D.C. Herzog, and K. Hinson.
1990. Plot size and shape in relation to soybean resistance for velvetbean
caterpillar (Lepidoptera:Noctuidae). J. Econ. Entomol. 83:2107-2110.
20. Funderburk, J.E., D.L. Wright, and I.D. Teare. 1990. Preplant tillage effects on
population dynamics of soybean insect pests. Crop Sci. 30:686-690.
21. Gamble, E.E. 1962. Gene effects in com (Zea mays L.) I.separation and relative
importance of gene effects for yield. Canadian J. Plant Sci. 42:339-348.
22. Gary, D.J., L. Lambert, and J.D. Ouzts. 1985. Evaluation of soybean plant
introductions for resistance to foliar feeding insects. J. MS Academy Sci. 30:67-
82.
23. Hallauer, A.R., and J.B. Miranda. 1981. Quantitative genetics in maize breeding.
Iowa State Univ. Press, Ames, IA.

97
24. Hamer, J.L. 1990. Soybean looper: biology and approaches for improved
management. Mississippi Cooperative Extension Service Information Sheet 1400.
25. Hart, S.V., J.W. Burton, and W.V. Campbell. 1988. Comparison of three
techniques to evaluate advanced breeding lines of soybean for leaf-feeding
resistance to com earworm (Lepidoptera:Noctuidae). J. Econ Entomol. 81:615-
620.
26. Hartley, G.G. 1990. Multicellular rearing methods for the beet armyworm,
soybean looper, and velvetbean caterpillar (Lepidoptera:Noctuidae). J. Entomol.
Sci. 25:336-340.
27. Hartwig, E.E., and J.M. Epps. 1967. Registration of Centennial soybeans. Crop
Sci. 17:979.
28. Hartwig, E.E., L. Lambert, and T.C. Kilen. 1990. Registration of Lamar
soybean. Crop Sci. 30:231.
29. Hartwig, E.E., S.G. Tumipseed, and T.C. Kilen. 1984. Registration of soybean
germplasm line D75-10169. Crop Sci. 24:214-215.
30. Hatchett, J.H., G.L. Beland, and T.C. Kilen. 1979. Identification of multiple
insect resistant soybean lines. Crop Sci. 19:557-559.
31. Hayman, B.I. 1958. The separation of epistatic from additive and dominance
variation in generation means. Heredity 12:371-390.
32. Hinson, K., R.A. Kinlock, H.A. Peacock, W.H. Chapman, and W.T. Scudder.
1981. Braxton soybean. Florida Agrie. Exp. Stn. Circular S-276.
33. Horsfall, J.G., and R.W. Barratt. 1945. An improved grading system for
measuring plant disease. Phytopathology 35:655(abstr.).
34. Jennings, C.W., W.A. Russell, W.D. Guthrie, and R.L. Grindeland. 1974.
Genetics of resistance in maize to second-brood European com borer. Iowa State
J. Research 48:267-280.
35. Joshi, J.M., and M. Nobakht. 1988. Evaluation of commercial soybean cultivars
and advance breeding lines for nonpreference to Heliothis zea. Soybean Genetics
Newsletter. 15:124-126.
36. Kilen, T.C., J.H. Hatchett, and E.E. Hartwig. 1977. Evaluation of early
generation soybeans for resistance to soybean looper. Crop Sci. 17:397-398.

98
37. Kilen, T.C., and L. Lambert. 1986. Evidence for different genes controlling
insect resistance in three soybean genotypes. Crop Sci. 26:869-871.
38. Kogan, M. 1982. Plant resistance in pest management, p. 93-134. In R.L.
Metcalf and W.H. Luckmann (ed.) Introduction to insect pest management. 2nd
ed. John Wiley and Sons, New York, NY.
39. Kogan, M. 1986. Natural chemicals in plant resistance to insects. Iowa State J.
Research 60:501-527.
40. Kogan, M., and D.E. Kuhlman. 1982. Soybean insects: identification and
management in Illinois. Illinois Agrie. Exp. Stn. Bull. 773.
41. Kogan, M., and S.G. Turnipseed. 1980. Soybean growth and assessment of
damage by arthropods, p.3-29. In M. Kogan and D.C. Herzog (eds.) Sampling
methods in soybean entomology. Springer, New York, NY.
42. Kraemer, M.E., M. Rangappa, P.S. Benepal, and T. Mebrahtu. 1988. Field
evaluation of soybeans for Mexican bean beetle resistance. I. Maturity groups VI,
VII, and VIII. Crop Sci. 28:497-499.
43. Lambert, L. 1984. An improved screen-cage design for use in plant and insect
research. Agron. J. 76:168-170.
44. Lambert, L., and J.L. Hamer. 1988. Evaluation of 10 soybean cultivars for
relative levels of damage by two insect species. Mississippi Agrie, and Forestry
Expt. Stn. Res. Rep. 13(5).
45. Lambert, L., and T.C. Kilen. 1984a. Influence of three soybean plant genotypes
and their F, intercrosses on the development of five insect species. J. Econ.
Entomol. 77:622-625.
46. Lambert, L., and T.C. Kilen. 1984b. Insect resistance factor in soybean Pi’s
229358 and 227687 demonstrated by grafting. Crop Sci. 24:163-165.
47. Lambert, L., and T.C. Kilen. 1984c. Multiple insect resistance in several soybean
genotypes. Crop Sci. 24:887-890.
48. Leonard, B.R., D.J. Boethel, A.N. Sparks, Jr., M.B. Layton, J.S. Mink, A.M.
Pavloff, E. Burris, and J.B. Graves. 1990. Variations in response of soybean
looper (Lepidoptera: Noctuidae) to selected insecticides in Louisiana. J. Econ.
Entomol. 83:27-34.

99
49. Leonard, W.H., H.O. Mann, and L. Powers. 1957. Partitioning method of
genetic analysis applied to plant-height inheritance in barley. Colo. Agrie. Exp.
Stn. Tech. Bull. 60:2-24.
50. Luckmann, W.H., and R.L. Metcalf. 1982. The pest management concept, p. 1-
28. In R.L. Metcalf and W.H. Luckmann (ed.) Introduction to insect pest
management. 2nd ed. John Wiley and Sons, New York, NY.
51. Mather, K., and J.L. Jinks. 1971. Biometrical genetics. 2nd ed. Chapman and
Hall, London.
52. McKenna, T., L. Lambert, J.D. Ouzts, and T.C. Kilen. 1988. Evaluation of wild
soybean, Glycine soja, for resistance to foliar feeding insects. J. MS Academy
Sci. 33:17-24.
53. Mebrahtu, T., W.J. Kenworthy, and T.C. Elden. 1988. Inorganic nutrient
analysis of leaf tissue from soybean lines screened for Mexican bean beetle
resistance. J. Entomol. Sci. 23:44-51.
54. Mebrahtu, T., W.J. Kenworthy, and T.C. Elden. 1990. Genetic study of
resistance to the Mexican bean beetle in soybean lines. J. Genect. and Breed.
44:7-12.
55. Meredith, W.R., Jr., and R.R. Bridge. 1972. Heterosis and gene action in cotton,
Gossypium hirsutum L. Crop Sci. 12:304-310.
56. Meredith, W.R., and M.L. Laster. 1975. Agronomic and genetic analysis of
tarnished plant bug tolerance in cotton. Crop Sci. 15:535-538.
57. Painter, R.H. 1968. Insect resistance in crop plants. The Univ. Press of Kansas,
Lawerence, KS.
58. Pedigo, L.P. 1989. Entomology and pest management. Macmillan, New York,
NY.
59. Plaut, J.L., and R.D. Berger. 1980. Development of Cercosporidium personatum
in three peanut canopy layers. Peanut Sci. 7:46-49.
60. Poehlman, J.M. 1987. Breeding field crops. 3rd ed. Van Nostrand Reinhold,
New York, NY.
61. Powers, L., and L.F. Locke. 1950. Partitioning method of genetic analysis
applied to quantitative characters of tomato crosses. USDA Tech. Bull. No. 998.

100
62. Rose, R.L., B.R. Leonard, T.C. Sparks, and J.B. Graves. 1990. Enhanced
metabolism and knockdown resistance in a field versus a laboratory strain of the
soybean looper (Lepidoptera: Noctuidae). J. Econ. Entomol. 83:672-677.
63. Rose, R.L., T.C. Sparks, and C.M. Smith. 1989. The influence of resistant
soybean (PI 227687) foliage and coumestrol on the metabolism of xenobiotics by
the soybean looper, Pseudoplusia includens (Walker). Pest. Biochem. and
Physiol. 34:17-26.
64. Rowan, G.B., H.R. Boerma, J.N. All, and J. Todd. 1991. Soybean cultivar
resistance to defoliating insects. Crop Sci. 31:678-682.
65. Rowe, K.E., and W.L. Alexander. 1980. Computations for estimating the genetic
parameters in joint-scaling tests. Crop Sci. 20:109-110.
66. Ruesink, W.G. 1980. Introduction to sampling theory, p 61-77. In M. Kogan and
D.C.Herzog (eds.) Sampling methods in soybean methodology. Springer, New
York, NY.
67. Sage, G.M.C., and M.J. de Isturiz. 1974. The inheritance of anther extrusion in
two spring wheat varieties. Theor. Appl. Genet. 45:126-133.
68. SAS Institute. 1990. SAS/STAT user’s guide. Vols. 1 and 2. SAS Inst., Inc.,
Cary, NC.
69. Schillinger, J.A. 1976. Host plant resistance to insects in soybeans, p.579-583.
In L.D. Hill (ed.) World soybean research. Interstate Printers and Publishers.
Danville, IL.
70. Scott, G.E., A.R. Hallauer, and F.F. Dicke. 1964. Types of gene action
conditioning resistance to European com borer leaf feeding. Crop Sci. 4:603-605.
71. Shokes, F.M., R.D. Berger, D.H. Smith, and J.M. Rasp. 1987. Reliability of
disease assessment procedures: a case study with late leafspot of peanut.
Oleagineux 42:245-251.
72. Sisson, V.A., P.A. Miller, W.V. Campbell, and J.W. Van Duyn. 1976. Evidence
of inheritance of resistance to the Mexican bean beetle in soybeans. Crop Sci.
16:835-837.
73. Smith, C.M. 1985. Expression, mechanisms, and chemistry of resistance in
soybean, Glycine max L. (Merr.) to the soybean looper, Psuedoplusia includens
(Walker). Insect Sci. Applic. 6:243-248.

101
74. Smith, C.M. 1989.Plant resistance to insects: a fundamental approach. John
Wiley and Sons, New York, NY.
75. Smith, K.J., and W. Huyser. 1987. World distribution and significance of
soybean. In J.R. Wilcox (ed.) Soybean: improvement, production, and uses. 2nd
ed. Agronomy 16:1-22.
76. Snedecor, G.W., and W.G. Cochran. 1989. Statistical methods. 8th. ed. Iowa
State University Press, Ames, IA.
77. Sokal, R.F., and F.J. Rohlf. 1981. Biometry. 2nd. ed. W.H. Freeman and
Company, New York, NY.
78. Southwood, T.R.E. 1978. Ecological methods with particular reference to the
study of insect populations. 2nd. ed. Chapman and Hall, New York, NY.
79. Sullivan, M.J. 1985. Resistance to insect defoliators. In R. Shibles (ed.) World
Research Conference III: Proceedings. Westview Press, Boulder, CO.
80. Talekar, N.S., H.R. Lee, and Suharsono. 1988. Resistance of soybean to four
defoliator species in Taiwan. J. Econ. Entomol. 81:1469-1473.
81. Tester, C.F. 1977. Constituents of soybean cultivars differing in insect resistance.
Phytochemistry. 16:1899-1901.
82. Tumipseed, S.G., and M.J. Sullivan. 1976. Plant resistance in soybean insect
management, p.549-560. In L.D. Hill (ed.) World Soybean Research. Interstate
Printers and Publishers. Danville, IL.
83. Van Duyn, J.W., S.G. Turnipseed, and J.D. Maxwell. 1971. Resistance in
soybeans to the Mexican bean beetle: I. Sources of resistance. Crop Sci. 11:572-
573.

BIOGRAPHICAL SKETCH
Michael Montgomery Kenty was bom on August 27, 1959 in San Antonio, Texas.
He completed his secondary education at Greenville High School, Greenville, Mississippi
in 1977. In 1977, he initiated his undergraduate studies at the University of Mississippi
in Botany. He received a Bachelor of Science in Agronomy at Mississippi State
University in 1983.
Prior to graduation he joined the staff of Sandoz Crop Protection Research, where
he conducted pesticide evaluations until September, 1986. In December, 1986, he joined
the staff of the USDA-ARS Soybean Production Research Unit at Stoneville, Mississippi
as an Agronomist. In December, 1988, he was awarded a Masters of Science in Natural
Science from Delta State University. In September, 1989, he took a leave of absence
from the USDA and entered the University of Florida. Currently, he is a candidate for
the degree of Doctor of Philosophy, Department of Agronomy, University of Florida.
102

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Kuell Hinson, Chair
Professor of Agronomy
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Kenneth H. Quesenberry
Professor of Agronomy
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Associate Professor of Agronomy
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Associate Professor of
Entomology and Nematology

This dissertation was submitted to the Graduate Faculty of the College of
Agriculture and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
April 1994
Dean, College of Agriculture
Dean, Graduate School

L.0
filo
i ^
.«.a*
UNIVERSITY OF FLORIDA
3 1262 08556 9639