 

 

 

THESEB

r)

1.

GAN STATE UNIVE IRS

IIIIIIIIIIIIIIIIIII III III lIIlIIlIIIIIII

31293 01565 9356

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LIBRARY I

Michigan State
University

 

 

 

This is to certify that the
thesis entitled
RELATIONSHIPS BETWEEN

SELECTED BIOLOGICAL AND MANAGEMENT ATTRIBUTES

OF MICHIGAN POTATO PRODUCTION SYSTEMS
presented by

Rebecca L. Gore

has been accepted towards fulﬁllment
of the requirements for

Master of Science degree in Entomology

    

Major professor

Date /Z /,7/?4

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

 

PLACE N RETURN BOX to roman this chockom from your record.
TO AVOID FINES return on or bdoro duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU is An N'finnottvo ActioNEquoi Opportunity instituion
WM!

 

RELATIONSHIPS BETWEEN
SELECTED BIOLOGICAL AND MANAGEMENT ATTRIBUTES
OF MICHIGAN POTATO PRODUCTION SYSTEMS
By

Rebecca L. Gore

A THESIS

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

MASTER OF SCIENCE

Department of Entomology

1996

ABSTRACT
RELATIONSHIPS BETWEEN
SELECTED BIOLOGICAL AND MANAGEMENT ATTRIBUTES
OF MICHIGAN POTATO PRODUCTION SYSTEMS
By

Rebecca L. Gore

Potato early die is an important economic disease of potato in Michigan. Many
management decisions impact symptom expression for this disease complex. This study
consists of two components. One part involved a survey that generated a series of multiple
regression models that demonstrated the relationships between certain management strategies
and expected potato yields. It was found that ﬁrst dividing the state into regions and
management systems into rotation, irrigation and chemigation parameters explained more of
the variability in expected yields (r2=0.7986) than any other model tested (other models were
based on farm size, nematicide usage and rotation scheme). The objective of the second
component was to distinguish responses to potato early die pathogens (Eramlmhm
W and Meﬂmillimn dahliae) among ten different potato cultivars (Red dale, Kennebec,
Superior, Russet bm'bank, Norkota russett, Hudson, Desiree, Rosa, Snowden and Atlantic).
Several types of analyses were used in the study including AN OVA, linear regression,
rankings and relative yields. Summarizing the data, it was found that Hudson, Russet
burbank, Snowden and Superior were most susceptible to potato early die, while Atlantic,

Norkota russct and Desiree were most resistant.

DEDICATION

With deepest admiration and respect
I would like to dedicate
my thesis
to my mentor and teacher

Dr. George W. Bird

iii

ACKNOWLEDGMENTS

I want to express my sincere gratitude to all of the members of my guidance
committee. Before I began my study in Nematology I listened to Jim Miller give a talk at a
regional entomology meeting about ethics in science, and it was for that reason I wanted him
on my graduate guidance committee and now is the time to thank him for his earlier advice
and his contribution to my graduate study.

When I ﬁrst worked in the Department of Entomology, I worked for Drs. Dean
Haynes, Stuart Gage, and George Bird and I enjoyed working for them and have learned the
value of sound experimental design, data presentation and dedication to the land grant
philosophy. I appreciate all the time you took with me both as an employee and as a student.

A special thanks goes to the other member of my committee, Dr. J. Roy Black from
Agrimltural Economics. I was ﬁrst introduced to Dr. Black when I began a project with him,
and listened to the importance he placed on regression analysis. It will not be hard for anyone
to seethat his appreciation forthat branch Ofstatistics has rubbed oﬁ‘on me. I want to thank
him for this and for reading several interpretations of the survey data.

Throughout my studies and employment at MSU I have relied on my colleague and
ﬁiend Fred Warner. I cannot begin to count the number of times we designed experiments
or discussed papers and tried to improve on experimental designs of others. Speaking of

counting, a special thanks goes to Dr. Carl Chen, who spent the better part of one summer

iv

helping quantify nematode and Verticillium counts. I appreciated all of his time and effort.
I also need to thank John Davenport both for his help in the initial potato survey and for
implermnting some rather daunting experimental designs, back in the days when my name was
Rebecca "Replication” Mather. Fmally, I would like to thank Mike Bemey for his help in the
ﬁnal days of this long journey in keeping me on track, as best anyone can.

On a personal note, I want to thank my family, and especially my son, Jacob, for

tolerating the amount of time and eﬁ‘ort I have put into school over the years.

TABLE OF CONTENTS

LITERATURE REVIEW ............................................... 1
INTRODUCTION ............................................... 1
THE POTATO SYSTEM ......................................... 1

Introduction .............................................. l
The Origin of the Potato ..................................... 2
Morphology, Growth and Development of the Potato Plant .......... 2
Whole Plant Physiology ..................................... 9
Conclusion .............................................. ll
POTATO EARLY-DIE DISEASE COMPLEX ........................ 12
Xenisillium dahliae ........................................ 12
Wench“ mm ..................................... 12
Modelling the Interaction between
W Wang and Winn: dahiiﬂﬁ .................. 13

vi

ANALYSIS OF AGRONOMIC AND SYSTEM DESIGN PARAMETERS
OF MICHIGAN POTATO PRODUCTION UTILIZING LINEAR

MODELLING TECHNIQUES .......................................... 18
INTRODUCTION .............................................. 18
OBJECTIVES ................................................. 18
MATERIALS AND METHODS ................................... 19
RESULTS .................................................... 20

State analysis ............................................ 20
Farm size model .......................................... 35
Region model ............................................ 44
Nematicide use model ...................................... 58
Rotation model ........................................... 70
DISCUSSION ................................................. 81
Singie-eﬁ‘ect (simple linear regression models) ................... 81
Multiple regression models .................................. 84
Summary ............................................... 86

ASSESSMENT OF INTRA— AND INTER- SQLANIJM W
CULTIVAR RESPONSES TO W8 BENEIBAHS

AND W DAHLIAE ....................................... 87
INTRODUCTION .............................................. 87
OBJECTIVES ................................................. 88
METHODOLOGY ............................................. 88

Inocuium preparation ...................................... 88
Jolly Road Microtile Study .................................. 89

vii

MSU Greenhouse Study .................................... 89

Montcalm County Potato Research Farm Field Study .............. 90

Statistical analysis ......................................... 90

RESULTS .................................................... 91
Intra-cultivar speciﬁc observations ............................ 92

Inter-cultivar speciﬁc ﬁndings ................................ 99
DISCUSSION ................................................ 103
BIBLIOGRAPHY ................................................... 126

viii

LIST OF TABLES

ANALYSIS OF AGRONOMIC AND SYSTEM DESIGN PARAMETERS
OF MICHIGAN POTATO PRODUCTION UTILIZING LINEAR

MODELLING TECHNIQUES
Table 1. The utilization of chemical inputs (lb/A) in Michigan potato

production in 1988 (nﬁO). .................................. 22
Table 2. A multiple regression model correlating potato yields to the

chemical inputs of nitrogen, phosphorus, potassium, and sulﬁtr. ...... 22
Table 3. Irrigation and rotation schedules in Michigan potato production in

1988 (n=40). ............................................ 26
Table 4. A multiple regression model correlating potato yields to

irrigation and rotation. ..................................... 26
Table 5. A multiple regression model correlating potato yields to

chemical inputs and management practices ...................... 26
Table 6. Portion of Michigan potato growers utilizing at-plant insecticides

and nematicides in 1988 ..................................... 29
Table 7. The utilization of at-plant nematicides for Michigan potato

production in 1988 (n=40). .................................. 29
Table 8. A multiple regression model correlating potato yields

to Temik and chemigation. .................................. 29
Table 9. A multiple regression model correlating expected yields to

chernicai inputs, management practices, and nematicide use

in 1988 Michigan potato production. .......................... 33
Table 10. The economics of Michigan potato production in 1988 ............. 33
Table 11. An estimate of total Michigan potato production

byfarmsizein 1988 ........................................ 37

ix

Table 12.

Table 13.

Table 14.

Table 15.

Table 16.

Table 17.

Table 18.

Table 19.

Table 20.

Table 21.

Table 22.

Table 23.

Table 24.

The utilization of chemical inputs (lb/A) by farm size in
Michigan potato production in 1988. .......................... 37

Irrigation and rotation schemes in Michigan potato production
by farm size in 1988 ........................................ 37

Fraction of Michigan potato growers utilizing at-plant insecticides
and nematicides by farm size in 1988. .......................... 40

The utilization of at-plant nematicides for Michigan potato
production by farm size in 1988. .............................. 40

A multiple regression model correlating potato yields to
rotation schedule, irrigation usage, amount of Temik
applied, and chemigation usage by farm size. .................... 43

A revised multiple regression model correlating potato
yields to rotation schedule, irrigation usage, and

chemigation usage by farm size. .............................. 45
The economics of Michigan potato production by

farm size in 1988. ......................................... 45
An estimate of total Michigan potato production

by farming region in 1988. .................................. 45
The utilization of chemical inputs (lb/A) by region

in Michigan potato production in 1988. ........................ 49

Irrigation and rotation schedules in Michigan potato
production by region in 1988. ................................ 49

Portion of Michigan potato growers utilizing at-plant
insecticides and nematicides by region in 1988. . . . .‘ ............... 52

The utilization of at—plant nematicides for Michigan
potato production by region in 1988. .......................... 52

A multiple regression model correlating potato yields
to rotation schedule, irrigation usage, amount of Temik
applied, and chemigation usage by farming region. ................ 55

Table 25.

Table 26.

Table 27.

Table 28.

Table 29.

Table 30.

Table 31.

Table 32.

Table 33.

Table 34.

Table 35.

Table 36.

Table 37.

A revised multiple regression model correlating potato
yields to rotation schedule, irrigation usage, and
chemigation usage by region. ................................ 57

The economics of Michigan potato production by region
in 1988. ................................................ 57

An estimate of total Michigan potato production by
nematicide use in 1988 ...................................... 60

Utilization of chemical inputs (lb/A) by nematicide
usage in Michigan potato production in 1988 ..................... 60

Irrigation and rotation schedules in Michigan potato
production by nematicide usage in 1988 ......................... 60

Fraction of Michigan potato growers utilizing at-plant
insecticides and nematicides by nematicide usage in 1988 ............ 64

The utilization of at-plant nematicides for Michigan
potato production by nematicide usage in 1988. .................. 64

A multiple regression model correlating potato yields
to pesticide input by nematicide usage for Michigan
potato growers. .......................................... 66

A multiple regression model correlating potato yields
to rotation schedule, irrigation usage, and pesticide
usage by nematicide use ..................................... 68

A revised multiple regression model correlating potato
yields to irrigation usage and chemigation inputs by
nematicide use. ........................................... 68

The economics of Michigan potato production
by nematicide use in 1988. .................................. 69

An estimate of total Michigan potato production
by rotation schedule in 1988. ................................ 72

The utilization of nutrient inputs (lb/A)

by rotation schedule in Michigan potato
production in 1988. ....................................... 72

xi

Table 38.

Table 39.

Table 40.

Table 41.

Table 42.

Table 43.

Table 44.

Table 45.

Portion of farms under irrigation in Michigan potato
production by rotation schedule in 1988. ....................... 73

Portion of Michigan potato growers utilizing
at-plant insecticides and nematicides by rotation
schedule in 1988. ......................................... 75

The utilization of at-plant nematicides for Michigan
potato production by rotation schedule in 1988. .................. 75

A multiple regression model correlating potato
yields to irrigation and pesticide scaling by
rotation schedule. ......................................... 78

A revised multiple regression model correlating
potato yields to irrigation and pesticide scaling by
rotation schedule. ......................................... 78

The economics of Michigan potato production
by rotation schedule in 1988. ................................ 80

A summary of the single eﬁ‘ect models for

potato yield prediction for the state as a whole,

or by farm size, region, nematicide use, or rotation

schedule (r’). ............................................ 83

A summary of the multiple regression models
for the state of MI, and by farm size, region, nematicide
usage, and rotation schedule. ................................ 85

ASSESSMENT OF INTRA- AND INTER- SQLANIIM W CULTIVAR
RESPONSES TO W8 RENEIBANS
AND W DAHIJAE

Table 1.

Three way analysis of variance p—values for variation
due to cultivar, treatment and replication for

Mancini: Remnant and Mertisillium dahliae
inoculations and diﬁ‘erent methods for the quantiﬁcation

onglannm Wm tuber yields at three experimental
sites. .................................................. 106

xii

Table 2. Number of Manchu; mm (PP) recovered at
harvest (1.0 g of root tissue + 100 cc soil) ﬁ'om four

treatments x 10 cultivars at Jolly Road, ﬁve treatment

x ﬁve cultivars for the Greenhouse study and ﬁve

treatments x ﬁve cultivars at the Montcalm Potato

Research Farm ........................................... 107

Table 3. Number of mm dahﬁae (VD) recovered at
harvest (propagules/g of root tissue) ﬁ'om four treatments

x 10 cultivars at Jolly Road, ﬁve treatment x ﬁve cultivars
forthe Greenhouse study and ﬁve treatments x ﬁve cultivars
at the Montcalm Potato Research Farm. ...................... 108

Table 4. Potato tuber yields (g/plant) ﬁ'om four treatments x 10 cultivars
at Jolly Road, ﬁve treatment x ﬁve cultivars for the Greenhouse
study and ﬁve treatments x ﬁve cultivars at the Montcalm
Potato Research Farm. .................................... 109

Table 5. Two way analysis of variance (treatment and replication) p—values
for source of variation due to treatment for each cultivar at the
three experimental sites (Jolly Road, the Greenhouse and

Montcalm Potato Research Farm). ........................... 1 10
Table 6. Ranges for ﬁnal counts of pathogens and joint interaction for
Jolly Road microtile site with 10 potato cultivars. ................ 111
Table 7. Ranges for ﬁnal counts of pathogens and joint interaction for
Greenhouse study for ﬁve potato cultivars. ..................... 112
Table 8. Ranges for ﬁnal counts of pathogens and joint interaction for
Montcalm Study with ﬁve potato cultivars. ..................... 112
Table 9. A series of linear regression tuber yield models [(tuber yield =

constant + PP + VD + PPND) and (tuber yield = constant +
PPND)] for each often cultivars utilizing the data from the Jolly
Road site (highlighted cultivars signify the best ﬁt equation). ‘ ....... 113

Table 10. A series of linear regression tuber yield models using the
natural log transformwon [(tuber yield = constant + ln(PP) +
ln(VD) + ln(PP/VD» and (tuber yield = constant + ln(PP/VD))]
for each often cultivars utilizing the data {Tom the Jolly
Road site (highlighted cultivars signify the best ﬁt model). ......... 114

xiii

Table 11.

Table 12.

Table 13.

Table 14.

Table 15.

Table 16.

Table 17.

Table 18.

Table 19.

Table 20.

Table 21.

A series of linear regression tuber yield models [(tuber yield =

constant + PP + VD + PPND) and (tuber yield = constant +

PPND)] for each of ﬁve cultivars utilizing the data from the

greenhouse study (highlighted cultivars represent best ﬁt models). . . . 115

A series of linear regression tuber yield models using the natural

log transformation [(tuber yield = constant + ln(PP) +

ln(VD) + ln(PPND)) and (tuber yield = constant + ln(PPND))]

for each of ﬁve cultivars utilizing the data from the greenhouse

study (highlighted cultivars signify the best ﬁt model). ............ 116

A series of linear regression tuber yield models [(tuber yield =

constant + PP + VD + PPND) and (tuber yield = constant +

PPND)] for each of ﬁve cultivars utilizing the data from the

Montcalm research site (highlighted cultivars signify the best

ﬁt equation). ............................................ 117

A series of linear regression tuber yield models using harvest

pathogen data (tuber yield = Constant + PP + VD) and the

natural log transformation [(tuber yield = constant + ln(PP) +

ln(VD))] for each of ﬁve cultivars utilizing the data from the

Montcalm research site (highlighted cultivars signify the best

ﬁt model). ............................................. 1 18

Intra- and inter-cultivar comparisons of relative tuber yield
for each cultivar at the three experimental sites (J Olly Road,
the Greenhouse and Montcalm Potato Research Farm). ........... 119

Intra- and inter-cultivar tuber yield ranking comparisons
for the three research sites. ................................. 120

Individual, intra— and inter-cultivar yield loss comparisons at the
Jolly Road microtile site ................................... 121

Individual, intra- and inter-cultivar yield loss comparisons for
the greenhouse study. ..................................... 122

Individual, intra- and inter-cultivar yield loss comparisons for the four
treatments vs. the control at the Montcalm research site ............ 122

Number of replications for each of the experiments (except the
yield loss experiments) ..................................... 123

Number of replications for the yield loss experiments. ............. 124

xiv

Table 22. Summary of comparative statistics among the three experiments
for the Brandenshus mm and mm dahliae treatment. ..... 125

LIST OF FIGURES

ANALYSIS OF AGRONOMIC AND SYSTEM DESIGN PARAMETERS
OF MICHIGAN POTATO PRODUCTION UTILIZING LINEAR

MODELLING TECHNIQUES
Figure 1. Correlation of phosphorus inputs to expected yields

in Michigan potato production in 1988 ......................... 23
Figure 2. Correlation of phosphorus inputs to expected yields

in Michigan potato production in 1988 ......................... 23
Figure 3. Correlation of potassium inputs to expected yields

in Michigan potato production in 1988 ........................ .24
Figure 4. The correlation of sulfur inputs to expected yields

in Michigan potato production in 1988 ......................... 24
Figure 5. Correlation of portion of irrigated land to expected yields in Michigan

potato production in 1988 .................................. 27
Figure 6. Correlation of rotation (years out of potato) to expected yields

in Michigan potato production in 1988 ......................... 27
Figure 7. Correlation of Temik 15G inputs to expected yields

in Michigan potato production in 1988 ......................... 30
Figure 8. Correlation of portion of acres chemigated to expected yields

in Michigan potato production In 1988 ......................... 30
Figure 9. Correlation of nematicide ranking with expected potato yields

in Michigan potato production in 1988 ......................... 31
Figure 10. Correlation of direct costs associated with planting and

expected yields in Michigan potato production in 1988 ............. 34
Figure 11. Correlation of tuber selling price to expected yields in

Michigan potato production in 1988 ........................... 36

xvi

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Correlation of direct costs with proﬁt for Michigan potato

production in 1988 ........................................ 36
Correlation of rotation (years out of potato) to expected yields

in Michigan potato production by ﬁrm size in 1988 ............... 38
Correlation of portion of irrigated land to expected yields in

Michigan potato production by ﬁrm size in 1988 ................. 38
Correlation of Temik 15G inputs to expected yields

in Michigan potato production by ﬁrm size in 1988 ............... 41
Correlation of portion of acres chemigated to expected yields

in Michigan potato production by ﬁrm size in 1988 ............... 41
Correlation of direct costs associated with planting and expected

yields in Michigan potato production by farnr size in 1988 .......... 46
Michigan potato growing regions as deﬁned in this study ........... 47

Correlation of rotation (years out of potato) to expected yields
in Michigan potato production by ﬁrming region in 1988 ........... 50

Correlation of portion of irrigated land to expected yields in
Michigan potato production by ﬁrming region in 1988 ............. 50

Correlation of Temik 15G inputs to expected yields in Michigan
potato production by ﬁrming region in 1988 ................... .53

Correlation of portion of acres chemigated to expected yields in
Michigan potato production by farming region in 1988 ............. 53

Correlation of direct costs associated with planting and expected
yields in Michigan potato production by ﬁrming region in 1988 ...... 59

Correlation of nitrogen inputs to expected yields
in Michigan potato production by nematicide usage in 1988 ......... 61

Correlation of rotation (years out of potato) to expected yields
in Michigan potato production by nematicide usage in 1988 ......... 61

Correlation of portion of irrigated land to expected yields
in Michigan potato production by nematicide usage in 1988 ......... 63

xvii

Figure 27. Correlation of Temik 15G inputs to expected yields in Michigan
potato production by nematicide usage in 1988 ................... 63

Figure 28. Correlation of direct costs associated with planting and expected
yields in Michigan potato production by nematicide usage in 1988 . . . . 71

Figure 29. Correlation of portion of irrigated land to expected yields in
Michigan potato production by years out of potato in 1988 ......... 71

Figure 30. Correlation of chemical nematicide usage and expected yields
in Michigan potato production by years out of potato in 1988 ........ 76

Figure 31. Correlation of direct costs associated with planting and expected

yields in Michigan potato production by years out of potato
in 1988 ................................................. 82

xviii

LITERATURE REVIEW

INTRODUCTION

The potato early-die disease complex is an important aspect of potato production
in many geographical locations. Understanding the nature of this disease complex is
essential for development of future management strategies and tactics, and a thorough
understanding of the potato plant and the production system is requisite to understanding
the disease complex. The literature review, therefore, is divided into sections on the

potato production system and the potato early-die causal components.

THE POTATO SYSTEM
Introduction

In 1980, it was estimated that potatoes were cultivated on 18,026,000 ha, yielding
225,676,000 tons of food and feed. Approximately 40% of the production was used for
human consumption and 31% for livestock (Burton 1989). The most common cultivated
potato in the United States is classiﬁed in the plant ﬁmily Solanaceae as Sglanum
mm (Hooker 1990). Propagation can be either ﬁ'om true seed or tubers with
tubers being the most common in the US. (Burton 1989).

The potato plant is a herbaceous dicot, with a C3 pathway (Dwelle 1985). It is

composed of ﬂowers, fruit, leaves, above-ground stems, below-ground stems, stolons,

2

four types of roots (basal, nodal, stolon and tuber) and tubers. Ifthe potato is grown from
seed it has a primary tap root, hypocotyl, cotyledons, and epicotyle (Hooker 1990).
Potatoes grown ﬁ'om tubers have adventitious root systems.
The Origin of the Potato

Torn Dillehay, an anthropologist at the University of Kentucky, discovered
peelings from cultivated potatoes in a swamp in Chile. He had them carbon-dated and
found that they were 10,000 years old (Hughes 1991). It is believed the potato originated
near Peru, and was an important source of food for the Incas. In the early 15705, Spanish
explorers brought the potato plant to Spain to use as food for prisoners and slaves. From
Spain it was transported to Italy. Europeans, at this time, considered the plant more of a
curiosity than a food. Gradually, during the 17th Century, the potato plant spread to
China, Japan, Aﬁica, India and Southeast Asia New Zealand obtained it in the 1700s.
The ﬁrst records of the potato in the US. indicated that it came with John Smith in 1620
(Salaman 1989). The most tragic demonstration of its growing dominance was, of course,
the Irish Potato Famine of 1845 caused by W infestans (late blight) (Salaman
1989).
Morphology, Growth and Development of the Potato Plant

The above ground potato plant consists of stems, leaves, ﬂowers and fruit. The
below ground part of the plant consists of roots, stolons, and tubers. Each part of the
plant has speciﬁc functions as well as functions connecting it to whole-plant processes.

W The ﬂowering and ﬁuit set include the structure of the

ﬂower, the development of the ﬂower, plant growth regulators and external ﬁctors.

3

Stmgture. The potato ﬂower can be white, yellow, purple or blue. It is complete,
having sepals, petals, stamens and carpels (Hayward 193 8). The pollen is wind-borne.
Self-fertilization is the norm (Hayward 193 8).

W- Development begins with primordia aligned to the stem that looks
like a dome-shaped enlargement. From this, the pedicel of the ﬂower is formed which is
composed of primarily vascular tissue. The strands divide into ﬁve parts to form the
sepals, then ﬁve petals are formed, then the corolla, stamen, and carpels. Finally, the
vascular bundles form the ovaries.

mm. It is hypothesized that gibberellins are responsible for
initiating the ﬂowering process in the potato species Solammr andigena (Burton 1989).

mm. It is believed that ﬂowering is also induced by a number of
ﬁctors including: light, temperature, water supply, humidity, nutrition, and seed tuber
condition. These ﬁctors vary from cultivar to cultivar. Burton (1989) showed that the
same variety can respond to light diﬁ‘erently if environmental conditions are changed. He
concludes that most species will ﬂower at 15-16 hours of light, which is the typical
schedule used by plant breeders (Burton 1989).

W. The ﬁ'uit is brown or purplish-green, measuring about oneohalf
inch in diameter, and containing 200 to 300 seeds (Haywood 1938). The seeds are
yellowish brown, srnail, ﬂat and oval or kidney-shaped. The fruit contains a single,
massive integument and a long micropyle. The mature integument has three layers: an
external layer of epidermal cells, and intermediate zone, and a digestive zone covering the

endospenn (Haywood 193 8). This digestive layer disappears as the seed mature

(Haywood 1938).

4

W. The leaf canopy includes the structure of the leaf, its development,
and the growth process.

1mm. Potato leaves are arranged spirally, usually counter-clockwise in
orientation, with the petiole ensheathed about one-third of the circumference of the stem
at the node (Haywood 193 8). The ﬁrst leaves of a plant originating ﬁom seed are simple
with later leaves compound (Haywood 193 8). The leaﬂets are typical dicots, with netted
venation. Young leaﬂets are densely pubescent. Leaves originating ﬁ'om seed pieces, can
also begin with simple leaves, but many of these plants begin with complex leaves
(Haywood 193 s).

W. Leaves develop at nodes on the stem. The primordia ﬁrst
appear as small leaf buttresses but rapidly change in appearance to a small stalk with a
visible blade. The terminal leaﬂet develops much ﬁster than lateral leaves. Leafhairs
appear dense when the leaf is small, but decrease in density when the leaf enlarges. The
number of hairs remain the same; however, since the leaf is increasing in size it is much
less dense.

Each leaﬂet is composed of epidermis (outer edge), the mesophyll, with two types
of parenchyma cells, the palisade (just below the epidermis), with the spongy parenchyma
beneath it. Chloroplasts are contained in the palisade region (Haywood 193 8). As the leaf
primordia matures, small groups of procambial cells are formed, these become
protoxylem. Then the phloem is diﬁ‘erentiated (Haywood 193 8).

W. Leaf growth occurs primarily at night, and is negatively
correlated to water potential. In fact, the rate of leaf growth stops when the water

potential of the plant is greater than -5 bars (Gander and Tanner 1976). Gander and

5

Tanner (1976) also showed that leaf area never recovers from water shortages, even if
only one irrigation period is missed. Total leaf area in a treatment that missed one
irrigation period was reduced to as much as 60% of the controlled, unstressed plant.

Wen. The root system includes its structure, both primary and lateral root
systems, and adventitious roots.

Wm. There are two types of root systems in potato plants, depending on
the origin of the plant. Plants grown from seeds have one taproot and a well developed
lateral root system; whereas, plants grown from seed pieces have adventitious roots
arising ﬁ'om young sprouts developing from buds of the seed piece. The root system is
ﬁirly shallow, if water is plentiful. It tends to go deeper when water is more scarce. The
majority of the roots have diameters less than 0.2 mm (64%) (Lesczynski and Tanner
1976). Burton (1989) calculated that the surﬁce area of the root system should be
approximately ﬁve times the leaf area. Along with water, the roots take up the minerals
needed for essential plant functions. These include sulﬁrr, phosphorus, magnesium,
calcium, potassium, and nitrogen.

Warn. The vascular system of the potato is protosteie in
arrangement. The inner most part of the root is the primary xylem, followed by
parenchyma, where the cambium is later found, then phloem, pericycle, endoderrnis with
Casparian strips, the cortex and epidermis (Hayward 193 8). Later roots appear quickly,
originating from pericycle cells, rupturing the endodermis.

Wm. Adventitious roots develop from the pericycle cells of the
stele in close proximity to a nodal plate, generally in groups of three (Haywood 193 8).

The root growth is similar to that of a primary root, except they are not diarch, but

6

sometimes have three or more separate regions of growth. Artschwager suggested that
enzymes secreted by the endodermis help the rootlet push through the cortex (Haywood).

The adventitious roots have secondary growth, although it is not as quick to occur
as it is in primary roots. Scondary xylem contains ﬁbers and parenchyma. Secondary
phloem consists of sieve tubes, companion cells and parenchyma. The primary phloem is
crushed in the growth process (Haywood 1938). The cortex increases in size by cell
enlargement of parenchyma. Often the epidermis and cortex break away in the ﬁnal stages
of development, leaving the endodermis exposed.

W. The potato plant has both above-ground and below-ground stems
called rhizomes or stolons. Many researchers have shown that above-ground stems can
produce stolons, and stolons can exhibit apical dominance under the correct environmental
conditions or when induced by a plant growth regulator.

W. The ﬁrst diﬂ‘erentiation of the apical meristem
is with cells that form the epidermis and the procambial ring that separates the pith from
the cortex. This ring eventually becomes the stele, in the mature stern. In the vascular
region, the protoxylern cells differentiate, then the inner phloem and outer phloem. While
this is occurring, the epidermis is also developing a cuticle and guard cells. Between the
vascular region and the epidermis is the cortex, with an outer layer of collenchyrna cells,
an inner layer of parenchyma cells, and bordered by the endodermis which has distinct
Casparian strips. Pericyclic ﬁbers are just inside the endodermis, and also within the

vascular bundles (Haywood 193 8).

7
As the stern matures, the six vascular bundles change in size, with three larger and
three smaller, which gives the stem a triangular or rectangular appearance. The bundles
appear cylindrical because of the development of the interﬁscicular cambium.
51W. The stolon develops very much like aerial lateral stems, with
the following exceptions. Externally, they have a hooked tip, scaled leaves, and grow
diageotropically (Booth 1963). Internally, the cells of the epidermis are single-layered, the
cortex has less collenchyma cells than the stem, and there is proportionately more phloem
than xylem in the stolen (Haywood 193 8). In the region where tuberizaiton occurs, the
endodernris with its accompanying Casparian strips is extremely pronounced.
WWW. Kumar and Wareing
(1971) have shown that any lateral bud can become either a leafy shoot or stolen
depending on the presence or absence of three growth regulators; auxin, gibberellin and
cytokinins. They elaborated on earlier experiments by Booth (1963), where he cut oﬁ' the
tops of plants and found stolons grew upward without an application of IAA, and
remained lateral, when IAA was applied to the cut surface in an eﬁ‘ort to maintain apical
dominance. Kumar and Wareing found this was also true when a stern cutting was used
instead of a whole plant. In addition, they manipulated stolons that were already
developed, by cutting oﬁ‘ the plant tops. They found that stolons exhibited apical
dominance when the top was cut oﬁ‘, but were unable to grow green leaves without the
addition of a kinetin in the presence of light.
In an errort to understand why typical potato plants develop leafy aerial laterals at
the top of the plant and stolons at the bottom, Kurnar and Wareing initiated a group of

experiments where they manipulated the three above mentioned growth regulators and

8

concluded it was the presence or absence of these regulators that controlled the
emergence of a stolen or leafy shoot. They concluded with the hypothesis that auxins and
gibberellins were produced in the above-ground area, and the cytokinins were produced by
the roots. The auxins remained at the top, and reacted in the upper buds, giving them
apical dominance, while gibberellins ﬂowed to the darker, moister areas beneath the soil.
The cytokinins in the roots were attracted to the parts of the plant that exhibited apical
dominance, and travelled upward through the stems, leaving the stolons to develop
beneath the soil.

Tum. Tuber development includes tuber initiation and growth, the development
process, and the role of growth regulators in the development process.

W. Tubers form ﬁrst on the stolen closest to the seed piece
(Wurr 1977, Plaisted 1958). These tubers also tend to be the largest (Plaisted 1958).
Tuber growth is largely due to cell division, Plaisted (1958) estimated that cell numbers
increased 500-fold as tubers enlarged, while cells increased in size only 10-fold. Within
the tuber, the pith experienced the ﬁster rate of increase as compared to the cortex
(Plaisted 1958). It is unusual that cell division continues for such an extended amount of
time.

W. The tuber is morphologically very similar to a thickened stem. The
ﬁrst sign of tuberization is the enlargement of the stolon tip. The ﬁrst region to grow is
the pith, with other regions attempting to adapt. The cortical cells become ﬁlled with
starch, the endodermis eventually disappears because it cannot form Casparian strips. The

cork cambium is active throughout the growth process. Vifrth the formation of the

9
periderrrr, stomata appear. As the tuber develops, buds are differentiated in the axils of
small scale leaves, which establish vascular connections to the stele (Haywood 193 8).

WWW- Although the exact growth
regulator that induces tuber initiation is unknown, it has been demonstrated that it reacts
to short-day cues (Chapman 195 8). Chapman showed that potato plants would initiate
tuberization alter a 14-day interval of a 9 hr light regime, if young leaves were present, but
would not if the only available leaves were mature. Hence it was concluded that stimuli
were present in active above-ground growing points, but could only be induced when
daylight dropped to 9 hr (Chapman 1958).

Whole Plant Physiology

Photosynthesis. Dwelle (1985) states that photosynthesis accounts for more than
90% of potato dry weight. The potato has a C3 photosynthetic pathway. The most
important consequence of the C, pathway is a net reduction of eﬁ‘ective gross assimilation
by about 40% because of the competitive oxidase activity of ribulose 1,5-biphosphate
carboxylase during photorespiration (Burton 1989).

Factors that inﬂuence the rate of CO2 assimilation include: leaf and canopy
structure, leaf area, chlorophyll content, tuber growth rates (translocation), growth
regulators and cultivar diﬁ‘erences. Environmental ﬁctors such as light and temperature
are also key (Dwelle 1985). i

In general, light interception increases linearly with the canopy leaf area index
(LAI), until LAI reaches about 2.5, then the rate decreases until an LAI of 4.0, at which
time about 95% of light is intercepted. This appears to be the maximum allowable

(Dwelle 1985). Dwelle cited a study conducted by Bremmer and Taha to illustrate the

1 0
importance of canopy longevity, stating that these researchers found a ”direct linear
relationship between tuber yield and number of days that the LAT is maintained at values
greater than 3.0." In addition, it has been shown that younger leaves have much greater
assimilation rate potential than older ones.

CO2 is absorbed by the leaf through stomates. Stomatal Openings are sensitive to
temperature. The warmer it is, the greater the opening. However, greater temperatures
also increase water loss through transpiration. Water loss causes the leaf surface to cool,
causing stomata openings to decrease. Light also eﬁ‘ects the stomata. The maximum
amount of light a potato plant is able to synthesis is 1200 uE/m2/s (Dwelle 1985).

Chlorophyll can also be a limiting ﬁctor in carbon assimilation. When leaf area is
increasing rapidly, its chloroplast development lags behind, and for a period of time
photosynthetic activity per leaf area can actually decrease (Dwelle 1985).

Translocation. Translocation is the process of getting the carbon to various sinks
(cg. tubers). The pathway is through the phloem. It is behaved that sucrose is the main
transport sugar (Burton 1989). Translocation rates average about 0.5 g/h for the whole
plant during photosynthetic periods, or about 0.05 to 0.1 g/hr/tuber (Burton 1989).

The inﬂuence of this demand on the photosynthetic process is illustrated by the
ﬁct that photosynthetic rates increase during tuber bulking and that individual leaf rates
increase when part of the canopy is either damaged (Colorado potato beetle feedings) or
removed mechanically (Dwelle 1985).

W. Transpiration is the process of water loss by the plant through
stomata openings. It is driven by leaf surface temperature. Water vapor exits the plant

through the stomata, reducing water in intercellular spaces. This water is replaced by

11

water ﬁ'om interﬁbrillar spaces in cell walls of leaf tissue, creating a water potential
gradient. This forces water up through the xylem, which in turn pushes water up ﬁ'om the
roots. Water enters roots through root hairs and mycorrhizal ﬁrngi, moves into the roots
because of the water potential gradient, through the interﬁbrillar spaces of the cortex,
through the endodermis, and into the vascular system (Burton 1989). The net effect of
transpiration is the reduction in leaf temperature. It can be stressful for the plant if there is
not enough water in the soil to replace the amount lost by the plant.

W. For a plant to grow optimally and produce high yields,
it has to maintain a relative equilibrium among these physiological processes. Growth
regulators are probably responsible for this balance. It is hypothesized that many of these
processes are regulated by gibberellin/auxin ratios, as well as cytokinins and ethylene
(Dwelle 1985). The exact stimuli are unknown.
Conclusion

The potato plant has several important phases of development. It is necessary to
establish a good canopy, adequate stolon production, and successful tuber initiation and
development. These processes are well synchronized and overlap. First, the above-
ground plant emerges, stolons develop, tubers initiate and bulking occurs. For the plant to
produce an adequate yield, these processes have to be well synchronized. For example, it
is important that stolons are ﬁirly long before tubers initiate or only a few tubers will
form. It is also important to note the shape of the tuber growth curve. It is a sigmoid
curve (Milthorpe 1963). However, the shape of the curve is variety dependent; it may
also be linear, developing at a constant rate; or discontinuous, developing and stopping,

and developing again.

12

POTATO EARLY-DIE DISEASE COMPLEX

Although many studies site a list of organisms associated with potato early die
(Martin et al. 1982, Rowe et al. 1987, Powelson 1985, Wheeler & Riedel 1994, Wheeler
el al. 1994), this study is exclusively referring to the presence of two pathogens
leniciilinm dahiiae and Brandenshus seam.
I! |° 'ii' I i l'

Minimal dahliae is a fungal pathogen of potato (Smith, 1968; Davis, 1985;
Nicot and Rouse, 1987). There are two species ofyerﬁcillium that eﬁ‘ect potato
production, M. dahliae and y. albmamun (Isaac and Harrison, 1968). The species most
prevalent in Michigan is y. dahliae due largely to Michigan’s climatic conditions. It has
been found that even this species, has a great amount of variability in potato yield loss due
to diﬁ‘erences between pathotypes of y. dahliae (Botseas and Rowe, 1994).
madman: Remnant

The genus Manchu: has at least 15 species that are pathogenic on potato
(Brodie 1984). The species, 2. mm: (Cobb, 1917) Filipjev & Schuurrnans-
Stekhoven, 1941, is one of the most pathogenic (Brodie, 1984). It is also the most
prevalent species in Michigan (Bird, 1981). The nematode is a migratory endoparasite
that remains vermiforrn throughout its life cycle. The second-stage juvenile emerges from
the egg and moves into a host root. All stages are infective and enter the root behind the
root cap by cutting an entrance with the stylet (Theme, 1961). Once inside of the root,
the nematode excretes substances that cause necrosis of cells.

Pathogenicity of potato was ﬁrst reported in 1938 by Hastings and Bosher. It was

later conﬁrmed by Oostenbrink (1954, 1956) and Dickerson et al. (1965). It was found

13

that nematode damages varied among the diﬁ‘erent cultivars (Burpee and Bloom, 1978;
Olthof, 1986).
Modelling the Interaction between [mm mm: and mm ﬂaming

Since the early 1980's it has been a goal of many nematologists and plant
pathologists to quantify the relationship between root lesion nematode and M. dahliae on
potato. The early work was done primary at Ohio State University, through the use of a
multi-year, two-locational, microplot series of experiments.

It was during this time period that the relationship between root lesion nematode
and M. dahliae was described as synergistic (Martin et al., 1982). An operative deﬁnition
of synergism is that the combined eﬁ‘ect of two pathogens is greater than the sum of their
individual eﬁ‘ects. From early potato early die research, it was learned that this eﬂ‘ect was
most easily demonstrated when the two pathogens were kept well below their individual
pathogenicity threshold levels (Martin et al., 1982).

The preliminary early die work at Ohio State consisted on an eﬁ‘ort to prove the
synergistic relationship of the two pathogens at low inoculum levels on potato (cv.
Superior). It consisted of a series of experiments with zero, low, medium, and high
nematode and Verticillium inoculations, both alone and in combination. In 1980, analysis
of variance and mean separations showed that low levels of nematode and Verticillium
tuber yield plots were similar to the control, while low nematode and low Verticillium
together in a plot were statistically diﬁ‘erent from the control (Martin et al., 1982).

These experiments were continued in 1981, 1982 and 1983 at two locations with
diﬁ‘erent soil types (Rowe et al., 1985). The experiments were done in Wooster, Ohio

with silt loam soil and in Ceieryville, Ohio with riﬂe peat soil. Mean separations were not

14

presented. Instead, a series of p-values associated with AN OVAs were published for four
levels of nematodes and three levels of the fungus. Results varied by both location and
year. In silt loam soils, low levels of both pathogens resulted in lower yields than the
controls, 50% of the time. However, in riﬂe peat soils low levels of both pathogens
yielded less than the controls all of the time. High levels of both pathogens yielded less
than the controls at both sites over all four years.

Even though the ANOVAs and presented data show a trend of lower yields in
treated plots versus controlled plots, there was no statistical evidence of synergism
presented. For an interaction to be synergistic, it had to be shown that crop loss from low
Verticillium alone plus low levels of nematodes alone, were statistically less than cr0p loss
when the two were combined in the same plot at the same levels that were in each pot
separately. This statistical test cannot be shown with analysis of variance. ANOVAs can
show that low 12. mm and low mm is different ﬁ'om low 2. ms and it is
different ﬁ'om low lenicillium, but it cannot show that it is diﬁ‘erent from the sum of the
eﬁ‘ect of low 2. mm and low Verticillium, which is the operative deﬁnition of
synergism.

The next phase of modelling potato crop loss came with regression analysis using
the same Ohio State data (F rancl et al., 1987 and 1990). Eleven regression models were
analyzed for each location and year individually, and than larger data sets were used that
analyzed each location across years (Francl et al., 1987). Both transformed natural log
and untranforrned pro-season pathogen counts were regressed with relative potato yields
and the best ﬁt models were presented. Adjusted r2 were signiﬁcant eight out of eleven

years, ranging from 0.28 to 0.97). In general, the best fit models included the natural log

1 5

of lenmnmm x R. m: as one of the parameters, which indicated that their
relationship was nonlinear. From a biological perspective, it may be interpreted that the
pathogens cause proportionally more damage at lower levels than they do at higher levels.

When Francl et al. combined years, a much smaller adjusted r2 was reported. One
location (silt loam soil type) had an r2 of0. 12, and the other (riﬂe peat) had an r2 of0.44.
When both locations were combined, the adjusted r2 = 0.20. These results indicate that
there were considerable variability in yield losses between years indicating that
environmental ﬁctors might be a major indicator of yield loss.

Further data were analyzed by Francl et al. In 1990. They used the previous Ohio
State data plus two years of additional data to ﬁnd better best ﬁt regression models.
Along with preseason nematode and Verticillium data, they used a canopy parameter that
improved the ﬁt of the combined data at both locations. The silt loam location (Wooster,
Ohio) increased its 1'1 value ﬁ'om 0.12 to 0.28%, for a relative yield model. At the peat soil
location, the r’ improved also. It went from 0.44 to 0.55. When true yields were
regressed instead of relative yields, the r2 were much improved. In Wooster (silt loam
site), the r2 was 0.48, and at Celeryville the r2 was 0.94. Both best ﬁt regression equations
were similar in that they had negative coefﬁcients for the canopy parameter alone
(potatoes were planted between microtiles some of the years), with a positive parameter
for an interaction parameter between canopy and the natural log of the Medium x P.
m: interaction term. In addition, both equations had a negative coeﬁcient for the
natural log of the interaction between the two pathogens separate ﬁ'om the canopy and a
negative coeﬂicient for the natural log of Verticillium alone.

Francl et al. (1990) used regression residuals to try to explain inter-season

16

variability with four years of the Ohio State data. They found that for two years the
residuals were generally positive (greater than zero) and the other two years the residuals
were negative, which could indicate environmental variability. They also correlated
temperature and moisture and found that there was a strong negative correlation between
late season warm weather and tuber yields.

Wheeler et al. (1991) used the Ohio State data to test another hypothesis Of yield
loss. Their regression model predicted yield loss on the basis ofyermillimn pre-plant
densities and used 2. menus as an indicator parameter. Two nonlinear regression
models were ﬁt, one in the presence of the nematode and the second in its absence. A
combined regression analysis for this model had an r2 of 0.52.

In 1994, Wheeler and Riedel published a paper that expanded the interaction study
to include 2. Smihnerj as well as P. ms. Microplots at Celeryville (riﬂe peat soil)
were used between 1986-1988. Both ANOVA and regression models were used to
analyze the data. Winn) eﬁ'ected potato yields in all three years. Manchu:
mm caused loss in two of the three years, with an interaction between P. penetms
and V. dahliae in those same years as well. Manchu: scribneri showed an eﬂ‘ect in one
year, but there was no interaction with y. dahliae. However, this paper did not attempt to
statistically demonstrate a synergistic interaction.

Chen (1995) attempted to quantify yield loss as either synergistic, additive or
antagonistic for treatments that contained root lesion nematode alone, the fungus alone,
and the results when both pathogen were introduced in the same treatment. He measured
yield loss and added the yield loss in the two individual pathogen treatments to get a

predicted value for joint interaction. Ifthe actual joint interaction was statistically greater

l 7
than this number the interaction was considered synergistic, if it was equal than it was
additive, and less it was considered antagonistic. The majority of the treatments (9) were

considered additive, with one synergistic and two antagonistic.

ANALYSIS OF AGRONOMIC AND SYSTEM DESIGN PARAMETERS OF
MICHIGAN POTATO PRODUCTION UTILIZING

LINEAR MODELLING TECHNIQUES

INTRODUCTION

Many factors inﬂuence variability in potato tuber yields. The literature documents
irrigation, crop rotation, and chemical inputs as potential causes of variability along with pest
occurrences. The impact of these management strategies can be analyzed using a series of

linear models.

OBJECTIVES
It is the objective of this study to determine the impact that selected agronomic
management practices and farming system structure have on potato tuber yields (cwt/A).
Several hypotheses were examined which entailed a series of linear regression models that
explained the variability in expected tuber yields among growers. The series of hypotheses
developed for this project include:
a. Variability in potato production, on a statewide basis, is explained by
diﬁ‘erences in chemical nutrient inputs, rotation schedule, irrigation, use of

Temik 15G or chemigation.

18

19

b. Variability in potato production can be stratiﬁed by ﬁrm size, with rotation
schedule, irrigation, use of Temik 15G and chemigation having the largest
impact on expected yields.

c. Variability in potato production can be stratiﬁed by region, with rotation
schedule, irrigation, use of Temik 15G and chemigation having the largest
impact on expected yields.

d. Variability in potato production can be stratiﬁed by nerrraticide use, with
rotation schedule and irrigation having the largest impact on expected yields.

e. Variability in potato production can be stratiﬁed by rotation schedule, with

irrigation and pesticide usage having the largest impact on expected yields.

MATERIALS AND METHODS

Forty Michigan potato growers were interviewed in 1988 by a representative of
Michigan State University Department of Agricultural Econorrrics and the Department of
Entomology Nematology Program. They were surveyed in regards to ﬁrming system
practices and associated variable costs. They were also asked to estimate potato yields for
the 1988 growing season The growers were selected at random from a list of potato farmers
supplied by the Michigan Potato Industry Commission. Twenty-four percent of Michigan's
43,500 acres of potato production were included in the study.

The data ﬁ'orn the survey were analyzed using a series of linear regression models to
determine the importance of agronomic practices on expected yields. The ﬁctors evaluated

included nitrogen, phosphorus, potassium and sulfur inputs; nematicide use; irrigation

20

practices and crop rotation schemes. These models were developed using the techniques of
linear algebra (Stapleton 1995).

The general linear model formulation is as follows:

Y = mx + b + e
In matrix form:

Y = [X] [Bi + e

where

Y = array of Y values

X = linear model

B = array of estimated coeﬁicients

e = array of error terms
The following matrix manipulations were completed to estimate the beta values:

[B] = [Xexrexer'

This provides an unbiased estimate of the beta parameters (Stapleton 1995).

The models were developed in a stepwise ﬁshion. The data were initially analyzed
for the state as a whole. Then separate models were developed based on ﬁrm size, potato
production region, nematicide use, and rotation schedule. Each factor was analyzed
individually, then a multiple regression model was developed. For each model, a series of

simple regression models were used to determine the relative importance of each ﬁctor.

RESULTS
W. Chemical nutrient inputs included nitrogen, phosphorus, sulfur and

potassium applications. Nitrogen inputs ranged ﬁom 49.6 to 345.2 lb/A with a mean of 181.9

2 1
lb/A for Michigan (T able 1). Nitrogen inputs did not correlate with expected yields. It was
found that the amount of nitrogen applied per acre was not a good indicator of potato yield
(r2=0.014, p=0.469, Figure 1).

Phosphorus inputs ranged from 0 to 288 lb/A, with a mean of 136.3 lb/A (Table 1).
Phosphorus inputs could only explain 0.2% of the variability in expected potato yields
(p=0.775, Figure 2).

Potassium inputs ranged ﬁom 0 to 425.6 lb/A, with a mean of 214.9 lb/A (Table 1).
Amount of potassium also did not correlate to potato yields (r2 =0.032, p=0.273, Figure 3).

Sulfur inputs ranged ﬁom 0 to 120 lb/A, with a mean of 10.5 lb/A These inputs
accounted for 0.5% of potato yield variability (p=0.661, Figure 4).

Generating a multiple regression model that hypothesized that all four of these
chemical inputs inﬂuenced expected tuber yield, it was found that together these inputs
explained only 5.5% of the variability in potato yields, which was not signiﬁcant (p=0.731).

Y = 212.09 + 0.123“‘XN -0.164"‘XP +0.198"'XK -0.349*Xs Equation 1
where:

Y = Expected Potato Yield

N = Nitrogen input

P = Phosphorus input

K = Potassium input

S = Sulfur input
The only ﬁctor that was signiﬁcant at the p505 level was the constant (Table 2).

The farming system practices of irrigation and rotation were also analyzed. It was

estimated that approximately 85% of the growers irrigate at least a portion of their potato

land, with 82% of total potato acreage irrigated in 1988. An average of 74% of each ﬁrm

22

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 1. The utilimtion of chemical nutrients (lb/A) in Michigan potato production in
1988 (n=40).
Chemical Minimum Maximum Mean Standard
input deviation
Men 49.6 345.2 181.9 71.4
Phorphorus 0.0 288.0 136.3 61.6
Potassium 0.0 425.6 214.9 80.5
Sulfur 0.0 120.0 10.5 28.6
Table 2. A multiple regression model correlating potato yields to the cherrrical inputs
of nitrogen, phosphorus, potassium, and sulfur.
Variable Coefﬁcient P-value
Constant 212.091 0.001
Men (lb/A) 0.123 0.584
Phosphorus (lb/A) -0.256 0.527
Potassium (lb/A) 0.197 0.323
Sulfur (lb/A) -0.349 0.526

 

 

 

Expected potato yield

23

 

 

 

 

 

500‘
400'
300' 4..
2“)“ El
El
2 El
E100 El E
v o r . t t
0 100 200 300 400

Nitrogen (lb/A)

Figure 1. The correlation of nitrogen inputs to expected yields in
Michigan potato production in 1988 (y=223.4+0.15x,

 

 

 

r2=0.014).
500
400 E El
1' ElIEl'Il'iiil iii
a 3°°‘_,_ a A, , :-
§ 200‘ E er a ”la
A § £E E
<11!)
3 El
.5.
O I j I
o 100 200 300

Phosphorus (lb/A)

Figure 2. Correlation of phosphorus inputs to expected yields in
Michigan potato production in 1988 (y=260.5-0.07, r2=0.002).

 

 

 

Potassium (lb/A)

Figure 3. Correlation of potassium inputs to expected
yields in Michigan potato production in 1988
(y=207.5+0.02x, r2=0.032).

 

 

 

 

500
400 E]
El
3 300
E a i
3’ ‘ 2m *—
8‘ rs
g“; 100
s
8
o . r a
50 100 150
Sulfur (lb/A)

Figure 4. The correlation of sulfur inputs to expected yields
in Michigan potato production in 1988
(y=253.4-0.23x, r2=0.005).

2 5
was irrigated in 1988. The average length of a rotation scheme was 1.5 years out of potato
(Table 3).

The portion of irrigated land explained a large portion of the variability in potato
yields (r2=0.437, Figure 5). When the state was taken as a whole, years out of potato had no
direct relationship to potato yields (r2 =0 .007, Figure 6).

Next, two separate multiple regression models were analyzed. One assessed the
impact of irrigation and rotation together, and the other included the chemical inputs as well.
It was formd that irrigation and rotation together explained 45.4% of the variability in potato
yields (Equation 2, Table 4). When chemical inputs were included, the model explained
51.9% of the variability in potato yields (Equation 3, Table 5). Both models were statistically
signiﬁcant (p<.001).

Y = 115.071 + 10.579‘XR + 162.584“Xl Equation 2
where:

Y = Expected potato yield

R = Number of years out of potato

I = Portion of land irrigated
When including both chemical inputs and ﬁrming system practices, the model is:

Y = 71.280 + 0.159"'XN -0.213*X,, +0.208‘XK -0.200*X3
+ 10.579"'Xll + 162.584"Xl Equation 3
where:

Y = Expected potato yield

N = Nitrogen input

P = Phosphorus input

K = Potassium input

S = Sulﬁlr input

R = Number of years out of potato
I = Portion of land irrigated

26

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3. Irrigation and rotation schedules in Michigan potato production in 1988
(n=40).
Practice Minimum Maximum Mean Standard
deviation
Irrigation (portion of 0.0 1.0 0.74 0.379
ﬁrms irrigated)
Rotation (years out 0.0 5.0 1.50 1.109
of potato)
Table 4. A multiple regression model correlating potato yields to irrigation and
rotation.
Variable Coeﬁcient P-value
Constant 1 15.071 0.000
Rotation
(years out of potato) 10.579 0.302
Irrigation (portion of
, farms irrigated) 29.576 0.000
Table 5. A multiple regression model correlating potato yields to chemical inputs and
management practices.
Variable Coeﬁcient P-value
Constant 71.280 0.163
; Nitrogen 0.159 0345
Phosphorus -0.213 0.266
Potassium 0.208 0.162
Sulfur 0200 0.622
Irrigation (portion of
ﬁrms irrigated) 165.590 0.000
Rotation
(years out of potato) 10.025 0.334

 

 

 

 

 

27

 

 

 

0.0 0.2 0.4 0.6 0.8 1.0 1.2
Portion of irrigated land

Figure 5. Correlation of portion of irrigated land to expected yields
in Michigan potato production in 1988
(y=132.4+160.6x, r2=0.437).

 

 

 

0 1 2 3 4 5 6

Years out of potato

Figure 6. Correlation of rotation (years out of potato) to expected yields
in Michigan potato production in 1988 (y=240.4+7.08x, r2 =0.007).

28

Although both models were statistically signiﬁcant, the only individual ﬁctor that was
signiﬁcant was the portion of land irrigated (Table 5).

Evaluation of nematicide usage was an important component of the survey. Fifty
percent of the growers used no nematicide. Thirty-seven percent used an at-plant nematicide
and 25% used either a ﬁlmigant or a chemigant (Table 6). Of the growers that used at-plant
nematicides, Temik 15G was utilized the most (Table 7). Since four or less growers used
Mocap IOG, Furadan or a fumigant, these chemicals were not utilized for the regression
models.

Tenrik 15G explained 17.2% of the variability in potato yields (p=0.008, Figure 7).
Chernigation explained 31.1% of the variability in yield (p<0.001, Figure 8). Then a
ranking was developed that gave numerical value to different types of nematicide inputs,
ranging from 0 for no inputs to 9 for at-plant nematicide, fumigant and chemigant. It was
found that this ranking did not correlate with expected yields, for the state as a whole (Figure
9, r’=0.000).

A multiple regression model, correlating potato yields to both Temik 156 and
chemigation increased the r1 to 0.365 (Table 8). The equation is as follows:

Y = 207.085 + 17.238"'X.r + 119.591"‘Xc Equation 4
where:

Y = Expected potato yield

T = Amount of Tenrik applied at-plant
C = 1 if chemigant applied; 0 otherwise

29

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 6. Portion of Michigan potato growers utihzrn' ' g at-plant insecticides and
nematicides in 1988.
Chemical input Mean Standard deviation
At-plant insecticide 0.825 0.385
At-plant nematicide 0.375 0.490
Fumigant 0.100 0.304
Ch 'gant 0.150 0.362
No nematicide 0.500 0.506
Table 7. The utilization of at-plant nematicides for Michigan potato production in 1988
(n=40).
Chemical input Min Max Mean Standard
deviation
Temik 15G 0.0 3.0 1.5 1.3
Mocap IOG 0.0 2.7 0.1 0.5
Furadan 0.0 3.0 0.1 0.5
Table 8. A multiple regression model correlating potato yields to Temik and
chenrigation.
Variable Coeﬁcient P-value
Constant 207.09 0.000
Temik 15G 17.24 0.085
Chemigation l 19.59 0.002

 

 

 

 

 

30

Expected potato yield

 

 

Temik 15G (lb/ailA)

Figure 7. Correlation of Temik 15G inputs to expected yields in
Michigan potato production in 1988
(y=207.6+28.82x, r2 =0.172).

 

WWﬁdﬂ

 

0 I I I l l l I
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Portion of acres chemigated

Figure 8. Correlation of portion of acres chemigated to expected yields
in Michigan potato production in 1988 (y=229.7+141.96x,
r2=0.311).

31

 

 

 

ta 151

El

6 ' s 10
Nematicide ranking‘

‘0 = no nematicide or at-plant insecticide, 1 = at-planting inSecticide, 2=at-
plant nematicide, 3=fumigant, 4=chemigant. Total sum ranges from 0 to 9,
with 9 being the application of at—plant nematicide, fumigant, and chemigant.

Figure 9. Correlation of nematicide ranking with expected potato yields in
Michigan potato production in 1988 (y=255.9-0.2x, r2=0.000).

32

Then Temik 15G and chemigation variables were added to the larger model. This
model included chemical inputs, management practices and nematicide inputs.
Y = 89.755 + 0.038"'XN -0.154"‘XP + 0.179"‘XK - 0.054’Xs
+ 14.722"'Xll + 147.663“‘Xl + 17.238"'XT
+ 119.591"'Xc Equation 5
where:

Y = Expected potato yield

N = Nitrogen input

P = Phosphorus input

K = Potassium input

S = Sulfur input

R = Number of years out of potato

I = Portion of land irrigated

T = Amount of Temik applied at-plant

C = 1 if chemigant applied; 0 otherwise

This ﬁlll model for the state of Michigan explained 63.9% of the variability in potato
yields (p<0.001). The two most important components of this model were portion of land
irrigated and use of chemigation (Table 9).

Since it is diﬁcult to quantify every aspect of potato production, the parameter of
expenses was chosen as an indirect indicator of collective inputs. Expenses included all costs
that were directly related to planting. For the state of Michigan expenses ranged ﬁom
$317.01 per acre to $965.29 per acre with a mean of $557.93 (Table 10). It was found that
expenses explained 14.7% of the variability in yields (p=0.014, Figure 10). Along with
expenses, return over direct costs were considered. A major component of this includes both

potato yields and selling price. Expected yields ranged ﬁom 75 to 430 cwt/A with an average

33

 

 

 

 

 

 

 

 

 

 

Table 9. . A multiple regression model correlating expected yields to chemical inputs,
management practices, and nematicide use in 1988 Michigan potato
production.

Variable Coeﬁcient P-value

Constant 89.75 0.057
Nitrogen 0.04 0.816
Phosphorus -0. 15 0.373
Potassium 0.18 0.185
Sulfur -0.05 0.882
Irrigation 147.66 0.000
(portion of ﬁrms

. irrigated)

Rotation 14.72 0.145
(years out of potato)
Tenrik 15G -5.84 0.575

. Chemigation 100.13 0.003

 

 

 

 

 

Table 10. The economics of Michigan potato production in 1988.

 

 

 

 

 

 

 

 

 

 

 

Economic Min Max Mean Standard
indicator deviation
Expected yield 75.00 430.00 251.00 92.03
Expected selling 4.00 11.00 6.23 ‘ 1.79
price
Expenses 317.01 965.29 557.93 174.74
Expected proﬁt -84.50 2927.88 969.81 632.80

 

34

 

Expected potato yield
ED

(cthA)
8
a
El

 

Expenses (SIA)

Figure 10. Correlation of direct costs associated with planting and
expected yields in Michigan potato production in 1988
(y=138.2+0.2x, r2=0.147).

35

of 251 chA (Table 10). Selling price ranged from $4.00 to $11.00/cwt, with an average of
$6.23/cwt for Michigan.

Selling price was correlated to expected yields and it was found that selling price
explained only 4.9% of the variability in yields (p=0.172, Figure 11). The ﬁnal correlation
was an attempt to relate expenses to return over direct costs, and it was found that they did
not correlate (r’<0.001, Figure 12).

W. Michigan ﬁrrn systems were divided into three size categories.
Small ﬁrms were classiﬁed as 50 acres or less, medium-sized ﬁrms as 50 to 250 acres, and
large ﬁrms were over 250 acres. Approximately 0.5% of total potato acreage consisted of
small ﬁrms, 30% medium-sized ﬁrms and 70% large ﬁrms (Table 11). There was no
signiﬁcant diﬁ‘erences in how these farming categories used nutrient chemical inputs (Table
12).

The number of years out of potato production was inversely proportionate to ﬁrm
size. Smaller ﬁrms tended to rotate more often than larger farmers (T able 13). Number of
years out of potato was conelated to expected yields for each ﬁrm size category. Years out
of potato was negatively correlated with expected yield for small ﬁrms (rz=0.550, n=4,
p=0.229), and positively correlated with potato yields for medium farms (r’=0.163, n=22,
p=0.062), and positively correlated for large ﬁrms (r’=0.097, n=14, p=0.279). When a full
model was analyzed, it was found that this farm sizelrotation model explained 28.65% of the
variability in expected potato yields (Figure 13).

The portion of irrigated land ranged ﬁom a mean of 0.5 on small ﬁrms to 0.8 on
large ﬁrms (Table 13). Portion of irrigated acres positively correlated to potato yields for all

three ﬁrm sizes (Figure 14). All i2 of 0.595 was not statistically signiﬁcant for small ﬁrms

36

 

 

5m I
El
m E E g
-u . El EEIEI El E1
.2 3m IN
.2
i 200 ' 3 En El El
A a El a El
< Im II
E? J E! El
V O r l I l I
2 4 6 8 10 12
Selling price (Slcwt)

Figure 11. Correlation of tuber selling price to expected yields
in Michigan potato production in 1988
(y=321.6-11.34x, r2=0.049).

 

 

 

SCXIOq E
'5' El
g 2WD' 8
u . ”a El
glee-M“ 1—
. $ 535 w .. '5'
B tilEl Eta I.
5 or
is -
“$000 I r I I
200 400 600 800 1000
Expenses($/A)

Figure 12. Correlation of direct costs with profit for Michigan potato
production in 1988 (y=941.97+0.04x, r2=0.000).

37

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 11. An estimate of total Michigan potato production by ﬁrm size in 1988.
Size of n Survey Percent of State
ﬁrm acres total estimate
acres (acres)
<=50 4 57 0.005 241.53
50-250 22 3034 0.295 12,855.93
>250 14 7175 0.699 30,402.54
Total 40 10,266 1.000 43,500.00
Table 12. The utilization of chemical inputs (lb/A) by ﬁrm size in Michigan potato
production in 1988.

Chemical input Small ﬁrms Medium ﬁrms Large ﬁrms
Nitrogen 164.8 171.7 202.3
Phorphorus 150.7 134.8 134.4
Potassium 207.8 213.4 219.3
Sulfur 0.0 17.6 2.4

Table 13. Irrigation and rotation schemes in Michigan potato production by ﬁrm size
in 1988.
Practice Small ﬁrms Medium Large
ﬁlms farms

Irrigation
(portion of ﬁrms 0.5 0.7 0.8
irrigateg
Rotation scheme
(years out ofpotato) 2.0 1.8 0.9

 

 

 

 

 

 

 

38

 

 

 

Full Model
500' r2=0.2865
400 l
.
E 9
a p‘ °
g zoo. . .
A ' a 5' Small farm
< a
31m 0 9 Mediumfarm
E D Largefarm
01 I r- 1 r r r
0 1 2 3 4 5 6

Number of years out of potato

Figure 13. Correlation of rotation (years out of potato) to expected yields
in Michigan potato production by farm size in 1988
(Small farm: y=244.6-44.17x, r2=0.550; Medium farm:
y=210.1+26.93x, r2=0.163; Large farm: y=235.0+50.0x, r2=0.097).

Full Model
500 ‘ r2=0.5572

 

 

 

 

 

0.0 0.2 0.4 0.6 0.8
Portion of land irrigated

Figure 14. Correlation of poortion of irrigated land to expected yields
in Michigan potato production by farm size in 1988
(Small farm: y=100.0+112.5x, r2=0.595; Medium farm:
y=165.1+144.96x, r2=0.320; Large farm: y=55.8+274.5x,

r2=0.628).

3 9
(n=4, p=0.229). However, both the medium farm (n=22, p=,0.006) and large farm (n=14,
p=0.001) regression models were statistically signiﬁcant. Testing all three regressions
simultaneously as a ﬁll model, it was found that collectively 55.72% of the variability in
potato yields could be explained by the ﬁrm size-irrigation model.

In general, small ﬁrms did not use nematicides (Table 14). When small ﬁrm growers
used Temik 156 they used it below nematicidal rates, as an at-plant insecticide (active
ingredient is less than 3 .O/acre, Table 15). Medium ﬁrms were the most reliant on at-plant
nematicides, with 45.5% of the medium—sized ﬁrm growers using this type of chemical input
(Table 14). While large ﬁrms tended to use furnigants and chemigants (50%) (Table 14).

Correlation of Temik lSG inputs with expected potato yields by ﬁrm size, indicated
that this regression was statistically signiﬁcant for large famrs (n=14, p=0.025); while not for
small ﬁrms (n=4, p=0.452) and medium ﬁrms (n=22, p=0.062). Testing all three regressions
simultaneously as a full model, it was found that collectively 36.31% of the variability in
potato yields could be explained by the ﬁrm size/Temik 156 input model (Figure 15). A
regression model based on chemigation was also statistically signiﬁcant for large farms (n=14,
p=0.003) and not for medium ﬁrms (n=22, p=0.20). Small ﬁrms were not correlated because
no small ﬁrm used chemigation. Therefore, the ﬁll model for farm size/chemigation
consisted of two regressions and explained 33.18% of the total variability in potato yields
(Figure 16).

A multiple regression model was utilized to assess the impact of the rotation schedule,
irrigation usage, amount ofTanik 15G applied, and the use ofa chemigant by ﬁrm size. The

following equation has twelve estimated values (n=40, df=32, k=12, s’=3664.5, r2=0.6892).

40

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 14. Fraction of Michigan potato growers utilizing at-plant insecticides and
nematicides by ﬁrm size in 1988.
Chemical input Small farms Medium Large ﬁrms
farms
At-plant insecticide 0.500 0.818 0.929
At-plant nematicide 0.000 0.455 0.357
Pro-plant 0.000 0.091 0.143
Fu 'gant
Pro-plant 0.000 0.045 0.357
Chemigant
No nematicide 1.000 0.500 0.3 57
Table 15. The utilization of at-plant nematicides for Michigan potato production by
ﬁrm size in 1988.
Chemical Small farms Medium farms Large ﬁrms
input
Temik 156 1.088 1.357 1.862
(ai/A)
Mocap IOG 0.000 0.191 0.000
(ti/A)
Furadan 0.000 0.154 0.000
(ai/A)

 

 

 

 

 

 

 

41

Full Model
r2 = 0.3631

 

 

 

 

 

Temik 156 (ai/A)

Figure 15. Correlation of Temik 15G inputs to expected yields in Michigan
potato production by farm size in 1988 (Small farm: y=196.2-
36.73x, r2=0.301; Medium farm: y=219.5+21.65x, r2=0.147;

Large farm: y=190.6+48.79x, r2=0.354).

Full Model
r2 = 0.3318

 

 

 

 

0 I I ﬁ I I 1 I

0.0 0.2 0.4 0.6 0.8 1.0 1.2
Portion of land chemigated

Figure 16. Correlation of portion of acres chemigated to expected yields in
Michigan potato production by farm size in 1988 (Medium farm:
(y=244.0+105.6x, r2=0.081; Large farm: y=228.9+147.1x, r2=0.546).

4 2
For small ﬁrms: Equation 6a
Y = 121.52 + 10.84‘X1R+ 131.81‘X1,- 45.93X1T
For medium farms: Equation 6b
Y = 121.52 + 1084*in + 114.87'X2, - 375sz + 76.85"'X2C
For large ﬁrms: Equation 6c
Y = 121.52 + 23.99"X3ll + 145.87*X3, + 3.83'X3, + 96.62"‘X3C

where:

X1 = ﬁrm 50 acres or less

X2 = ﬁrm greater than 50, less than 250 acres

X3 = ﬁrm 250 acres or greater

Y = Expected potato yield

R = Number of years out of potato

I = Portion of land irrigated

T = Amount of Temik lSG applied (ai/A)

C = 1 if chemigant applied, 0 otherwise

Of these twelve indicators, six were signiﬁcant at the p<0.05 level (T able 16). A
smaller model was developed using these six estimators; portion of land irrigated for each of
the different ﬁrm sizes, years out of potato for medium sized ﬁrms and the indicator if
chemigant was applied for large sized ﬁrms. This model included six parameters (n=40,
df=34, k=6, s’=3429.6, rz=0.6468):
For small ﬁrms: Equation 7a

Y = 117.25 + 95.24X1l

For medium ﬁrms: Equation 7b

Y = 117.25 + 23.86x2R + 117.84X2,

43

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 16. A multiple regression model correlating potato yields to rotation schedule,
irrigation usage, amount of Temik applied, and chemigation usage by ﬁrm
size.

Coeﬁcient Beta Estimator P-value
Constant BO 121.52 0.000
Rotation R1 ' 10.84 ' 0.623
”mm" R2 23.99 0.046

R3 1.90 0.959
Irrigation 11 131.81 0.024
12 114.87 0.006
13 145.87 0.022
Temik T1 1 -45.93 0.174
T2 -3.75 0.756
T3 3.83 0.881
Chemigation C2 76.85 0.242
C3 96.62 0.019

 

 

 

 

 

 

4 4
For large ﬁrms: Equation 7c
Y = 117.25 + 160.73X3, + 98.01X3C
where:

Y = Expected potato yield

R = Number of years out of potato

I = Portion of land irrigated

C = 1 if chemigant applied, 0 otherwise
The revised model provides a good estimate of yield based on halfof the estimators. All six
estimators were signiﬁcant at the p<0. 10 level, and ﬁve of the six were signiﬁcant at the
p<0.05 level (Table 17).

After analyzing inputs, it is still necessary to examine the economics of potato
production by ﬁrm size. Large farms have more return over direct costs than medium and
small ﬁrms (Table 18). Expected yields ranged ﬁ'om 156.25 on small ﬁrms to 281.43 cwt/A
for large ﬁrms. Although smaller ﬁrms tended to receive a better selling price (a mean of
$7.62), they still average much less return over direct costs (Table 18).

Correlation of expenses (SIA) to return over direct cost (S/A) indicated that small
ﬁrms tend to lose more money, the more they spend on system inputs (n=4, p=0.510).
Medium ﬁrms lose somewhat less than small ﬁrms (n=22, p=0. 159), while large ﬁrms tend
to increase their proﬁts with increased expenditures (n=l6, p=0.131), (Figure 17).

W Michigan was divided into six regions; the Upper Peninsula (UP),
northern Lower Peninsula (N), West Central (WC), East Central (EC), Southwest (SW) and
Southeast (SE) (Figure 18). The largest potato producing regions are WC and EC Michigan
(T able 19). Since only two growers were surveyed in the SE, this region was not statistically

analyzed. Statistical results for the UP and the SW are also very limited in their applicability,

45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 17. A revised multiple regression model correlating potato yields to rotation
schedule, irrigation usage, and chemigation usage by ﬁrm size.
Coeﬁcient Beta Estimator P-value
Constant BO 1 17.25 0.000
Rotation schedule R2 23.86 0.027
Irrigation 11 95.24 0.050
12 1 17.84 0.000
13 160.73 0.000
, Chemigation C3 98.01 0.011
Table 18. The economics of Michigan potato production by ﬁrm size in 1988.
Economic indicator Small ﬁrms Medium ﬁrms Large ﬁrms
Expected yield 156.25 248.86 281.43
Expected selling price 7.62 5.95 6.26
Expenses 590.98 536.77 581.74
Expected proﬁt 472.14 906.99 1210.70
Table 19. An estimate of total Michigan potato production by ﬁrming region in 1988.
Region 11 Total acres Survey acres Percent
surveyed
UP 3 3500 260 7.43
North 8 7600 2145 28.22
West Central 12 12,100 3489 28.83 '
East Central 11 12,800 2205 17.23
Southwest 4 3250 467 14.37
Southeast 2 4250 1700 40.00
TOTAL 40 43,500 10,266 23.60

 

 

 

 

 

 

 

 

3WD '
g
3
§
8
8
ﬂ
Figure 17

 

46

Full Model
r2 = .2501

 

 

 

 

Expenses (S/A)

Correlation of direct costs associated with planting and

ed yields in Michigan potato production by farm size
in 1988 (Small farm: y=1113.8=1.09x, r2==0.240; Medium
farm: y=1374.8-0.87x, r2==0.097; Large farm: y=94.9+1.92x,
r2 =0.179).

  

47

Upper Penninsula
Northern Region
West Central Region

V//////. East Central Region
Southwest Region

EZIEII

Southeast Region

Figure 18. Michigan potato growing regions,

48

due the small sample sizes (three and four growers, respectively). The six regions did not
diﬁ'er signiﬁcantly regarding utilintion of nitrogen, phosphorus, potassium, and sulfur (Table
20).

Crop rotation schemes ranged from an average of one year out of potato in the SW
to two years out in the UP (Table 21). Correlation of rotation schemes to expected potato
yields in the UP, N, and NC regions indicated that the best ﬁt produced a negative correlation
for all three of these regions. Years out of potato explained about 25% of the variability in
both the UP and the N (p=0.667 and p=0.21, respectively). In the WC region, rotation did
not eﬁ‘ect yield (r’=0.000, p=0.981). However, number of years out of potato correlated
positively in the EC and explained about 14.3% of the variability in yield (p=0.251, Figure
19). When both number of years out of potato and region were correlated together with
expected yield, 53.07% of the total variability in expected yields were explained.

Irrigation ranged from 0.5 in the UP and EC to 1.0 in the WC and SW, where 0.5 is
an average of one-half of the acreage is irrigated and 1.0 means that total acreage is irrigated.
Irrigation correlated positively in four of the regions (UP, N, WC, and EC), and negatively
in the SW (Figure 20). The percent of variability in potato yields ranged from 11.9% in the
WC region to 65.6% in the North; with the full model (correlating region and portion of
irrigated land sinmltaneously with expected yield) explaining 71.9% of the variability (Figure
20). The regressions were statistically signiﬁcant for the N (p=0.015) and EC (p=0.009), but
not for the UP (p=0.667), WC (p=0.272) or SW (p=0.580).

There was no diﬁ‘erence among the regions in at-plant insecticide use with portion of
growers using these insecticides ranging ﬁ'om two-thirds of the growers in the UP to all of

the growers surveyed in the WC and SE regions. The range was greater for at-plant

49

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 20. The utilization of chemical inputs (lb/A) by region in Michigan potato
production in 1988.
, R__egion Nitrogen Phosphorus Potassium Sulfur
UP 157.0 181.2 229.4 0.0
North 217.0 144.3 181.3 15.0
West Central 187.4 108.6 244.7 10.1
East Central 156.8 117.5 230.5 9.8
Southwest 183.9 190.6 177.6 18.0
Southeast 180.5 198.0 138.0 0.0
Table 21. Ilrgisgsation and rotation schedules in Michigan potato production by region in
Irrigation Rotation
Mon (portion of ﬁrms irrigated) (years out of potato)
UP 0.5 2.00
North 0.6 1.75
West Central 1.0 1.75
East Central 0.5 1.09
Southwest 1.0 1.00
Southeast 0.7 1.50

 

 

 

 

 

 

50
Full Model

T 0 r2 = 0.5307

 

 

2%

WC

 

 

 

 

OﬂOE

 

1 1 l I

0 1 2 3 4 5 6
Number of years out of potato

Figure 19. Correlation of rotation (years out of potato) to expected yields
in Michigan potato production by farming region in 1988
(UP: y=300.0-12.5x, r2=0.250; N: y=336.3—64.3x, r2=0.256; WC:
(y=332.9-0.45x, r2=r2=0.000).

 

 

 

 

 

500 . Full Model
. r2 = 0.7191
400 " :
'6 ,
1: 300 ‘ , 4o
g. o . ‘ 8 UP
.8 2m . B/‘ o N
g. . , ° ° 3 . '1 WC
2 ' ° EC
3100 _ I SW
g ,
O I I I f I I I
0.0 0.2 0.4 0.6 0.8 1.0 1.2

Portion of land irrigated

Figure 20. Correlation of portion of irrigated land to expected yields
in Michigan potato production by farming region in 1988
(UP: y=262.5+25.0x, r2=0.250; N: y=70.1+245.8x, r2=0.656;
WC: y=-228.8+560.5x, r2=0.119; EC: y=140.5+101.3x, r2=0.549;
SW: y=1075.0-850.0x, r2=0.l76).

5 1
nematicides, with zero growers using them in the UP, and 75% of the growers using them in
the WC region. Fumigants and chemigants were only used by growers in the N and WC
regions (Table 22).

Temik 15Gwasused the most predominately in the WC region. The amount ofactive
ingredient per acre ranged from 0.00 in the UP to 2.613 1b ai/A in the WC region (Table 23).
Correlating Temik 15G usage and expected potato yields, it was found that Temik ISG
correlated positively in the N, WC, and EC regions; and negatively in the SW. However,
none ofthe regressions were statistically signiﬁcant (N: p=0.396; wc: p=0.284; EC: p=0.950;
SW: p=0.692). When both region and Temik ISG usage were correlated simultaneously with
expected yield, the model explained much more of the variability in potato yields (r2=0.462)
(Figure 21).

In the N and WC region, chemigants were also used by some of the growers. Thirty-
seven percent of the growers surveyed used a chemigant in the N region, while 25% of the
growers used them in 8 WC region (Table 22). Chernigant use correlated extremely well with
potato yields in the N region (r"=0.926, p<0.001). However, the correlation was not
statistically signiﬁcant in the WC region (13=0.160, p=0. 197). Correlating chemigant use and
region simultaneously, it was found that for the full model 83.57% of the variability in
expected yields was explained by the use of a chemigant (Figure 22). i

A multiple regression model was also utilized to explain variability in expected potato
yield. The variables of years out of potato, portion of land irrigated, the amount active
ingredient of Temik 156 utilized, and chemigant usage were regressed with expected potato

yields. This model included 16 parameters (n=38, s7=2232, F16, r2=0.815).

52

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

83 ooo.o 83 aw:
ooo.o ES 82 3%
980 02.0 $3 seﬁlnzslmm
omno ooo.o Sod 35:00 803
ooo.o ooo.o on“: 5.82
ooo.o ooo.o ooo.o .5
5mm memo 5E
Sosa 02 9822 o: cases 5&2
.39 E 8&0. 3 530.605 938 ﬁmzoaz 8o 0020395: E298 oo nouns—u: 2F .8 030,—.
ooo.o ooo.o ooo.o oOm .o ooo._ “meoﬁzom
oOmo ooo.o ooo.o cad omho um0>>5=om
ooo.o ooo.o ooo.o Rod and 35:00 “mam
ooo.o 036 03.0 omho ooo._ 35:00 603
oomo memo m2 .o ommo omho 552
ooo._ ooo.o ooo.o ooo.o hood 5
“53:25 Emmi—am 0220080: 02602:
0203050: 02 :8305 803.05 “Suva E292 =o_w0.m

 

 

 

 

 

 

.32 E .8300 3 0.0203080: 23 002030005 “53-3 9.3%... 9.038» 888 Sumac: me 5:00.—

 

.NN 030B

 

53

- Full Model
1 : r2 = 0.4621

 

 

 

 

 

OﬂOB
Pi

 

 

 

 

 

Temik 156 (ailA)

Figure 21. Correlation of Temik ISG inputs to expected yields in Michigan
potato production by farming region in 1988 (N: y=153.9+38.3x,
r2=0.112; WC: y=279.2+20.2x, r2=0.114; EC: y=193.9+1.7x, r2=0.000;
SW: y=254.8-12.5x, r2=0.095).

 

 

 

 

Full Model
500 =0.8357
400
'6
300
8.
.8
g. 200
is...
t
3
O I I I I 1 I I
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Portion of land chemigated

Figure 22. Correlation of portion of acres chemigated to expected yields in
Michigan potato production by farming region in 1988
(N: y=132.0+244.7x, r2=0.926; WC: y=320.6+46.1x, r2=0.160).

54

For the Upper Peninsula: Equation 88

Y = 151.49 + 37‘X1R + 99.01‘XlI

For the Northern Region: Equation 8b

Y = 151.49 - 19.97’X2R+ 33. 15*X21+ 3.88*X2T + 209.66*X2C

For the West Central Region: Equation 80

Y = 151.49+ 0.85‘X3R + 142.57*X31 + 10.61'X3T + 40.71‘X3C

For the East Central Region: Equation 8d

Y = 151.49 + 12.56*X4R + 85.74‘X4I - 29.15‘X4T

For the Southwestern Region: Equation 8e

Y = 151.49 + 100.85‘X51 - 10.26‘X5T

X1 = Upper Peninsula

X2 = Northern Region

X3 = West Central Region
X4 = East Central Region
X5 = Southwestern Region

Y = Expected potato yield

R = Number of years out of potato

I = Portion of land irrigated

T = Amount of Temik lSG applied (ai/A)
C = 1 if chemigant applied, 0 otherwise

Of the sixteen indicators, seven were signiﬁcantly different from zero at the p<0. 10

level (Table 24). These were used to build a smaller model based on region (n=3 8, df=31,

1r=7, s’=2064.6, H.799).

For the Upper Peninsula: Equation 9a

Y = 135.58 + 42.30‘X1R + 109.6*XlI

55

Table 24. A multiple regression model correlating potato yields to rotation schedule,
irrigation usage, amount of Temik applied, and chemigation usage by farming

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

region.

Coeﬂicient Beta Estimator P-value
Constant B0 151.49 0.000
Rotation R1 37.00 0.023
”mu!" R2 -1997 0.250

R3 0.85 0.963
R4 12.56 0.247
Irrigation 11 99.01 0.051
12 33. 15 0.677
I3 142.57 0.017
14 85.74 0.014
I5 100.85 0.015
Temik T2 3 .88 0.886
T3 10.61 0.580
T4 -29. 15 0.23 1
T5 -10.26 0.566
Chemigation C2 209.66 0.000
C3 40.71 0.238

 

 

 

 

 

 

56

For the Northern Region: Equation 9b
Y = 135.58 + 241.07*X2C
For the West Central Region: Equation 9c
Y = 135.58 + 198.68‘X31
For the East Central Region: Equation 9d
Y = 135.58 + 107.06’X4I
For the Southwestern Region: Equation 9e
Y = 135.58 + 102.79"X51
where:
X1 = Upper Peninsula
X2 = Northern Region
X3 = West Central Region

X4 = East Central Region
X5 = Southwestern Region

Y = Expected potato yield
R = Number of years out of potato
I = Portion of land irrigated
C = 1 if chemigant applied, 0 otherwise
All seven of these parameters were signiﬁcant at the p_<_0.031 level (Table 25).

Alter reviewing the impact of speciﬁc inputs individually and together, it is important
to look at the economies of the production system by region Average expected yields ranged
ﬁom 175 cwt/A in the SE to 332 cwt/A in the WC region. Selling prices ranged ﬁ'om 5.50
in the SW, to 8.09 in the N region. Expenses also varied among regions from $366.85/A in
the UP to $680.53/A in the WC region. Expected proﬁt ranged from $588.12/A in the EC
region to $1549.82/A in the UP (n=3, Table 26). Correlating expenses to return over direct

cost (proﬁt), best ﬁt regressions indicated positive correlations in the UP and N regions

57

 

 

 

 

 

 

 

 

Table 25. A revised multiple regression model correlating potato yields to rotation
schedule, irrigation usage, and chemigation usage by region.
Coeﬁcient Beta Estimator P-value

Constant B0 135.58 0.000

Rotation R1 42.30 0.009

schedule

Irrigation 11 109.60 0.031
13 198.68 0.000
14 107.06 0.001
15 102.79 0.001

Chemigation C2 241.07 0.000

 

 

 

 

 

 

Table 26. The economics of Michigan potato production by region in 1988.

 

 

 

 

 

 

 

 

Expected Expected Expected
Region yield selling price Expenses proﬁt
(cm/Q (Slew!) (WA) (S/A)
UP 275.00 7.00 366.85 1549.82
North 223.75 8.09 544.25 1173.45
West Central 332.08 5.56 680.53 1168.53
East Central 194.55 5.51 484.97 588.12
Southwest 237.50 5.50 613.39 680.37
Southeast 175.00 7.00 454.00 771.00

 

 

 

 

 

 

58

(r’=0.847, p=0.256; r7=0.009, p=0.825, respectively), and negative correlations in the WC,
EC, and SW regions (r’=0.023, p=0.638; r2=0.062, p=0.458; rz=0.48, p=0.782, respectively).
Hence, there is no statistically signiﬁcant correlation of expenses to proﬁt in any of the
regions tested (Figure 23).

W. Of the forty growers surveyed, 50% of them used either a
non-fumigant, ﬁmigant, and/or chemigant nematicide as part of their management practice.
These growers comprised about two-thirds of the total acreage planted to potato in the
surveyed area (Table 27). Two or more nematicides were used in several cases.

There were no signiﬁcant differences with how these two groups used nutrient inputs.
However, there was a greater absolute difference between these two groups in mean nitrogen
inputs, so nitrogen was incorporated into the larger regression model (Table 28). Correlating
nitrogen inputs to expected potato yield by nematicide usage, it was found that nitrogen
inputs did not explain yield variability (r2=0.000, p=0.947 for growers using nematicides;
r’=0.032, p=0.454 for growers not using nematicides). However, when correlating both
nematicide use and nitrogen inputs with expected potato yields simultaneously, it was found
that the full model explained 44% of the variability in yields (Figure 24).

Growers that used nematicides tended to rotate for longer periods than those that did
not use nematicides (Table 29). Years out of potato did not explain expected yield variability
(r2=0.014, p=0.625 for growers using nematicides; r2=0.003, p=0.807 for growers not using
nematicides). However, when both nematicide usage and number of years out of potato were
correlated together, the full model explained 43.74% of the variability in potato yields (Figure

25).

 

 

 

 

 

 

 

 

 

 

 

30001 e
2000‘ . ll Full Model
.8 r2=0.3102
§ . on" n _.. ,
Elmo-W ,a .T 13"
O __

E °° I 8 1‘ " WC
A 0" o 0 EC
5% I sw
are. ,
'1“ I j I I

200 400 600 8m 1000
Expenses (SIA)

Figure 23. Correlation of direct costs associated with planting and
expected yields in Michigan potato production by
farming region in 1988 (UP: y=-2544.4+11.2x, r2=0.847;
N: y=891.7+0.5x, r2=0.009; WC: y=1455.4-0.4x, r2=-0.023;
EC: y$00.2-0.6x, r2=0.062; SW: y=987.6—0.5x, r2=0.048).

60

Table 27 . An estimate of total Michigan potato production by nematicide use in 1988.

 

 

 

 

 

 

Treatment Growers Survey Percent of State
(11) acres acreage (%) estimate
(acres)
Nematicide used 20 6919 67.39 29314.65
No nematicide 20 3347 32.60 14181.00

 

 

 

 

Table 28. Utilization of chemical inputs (lb/A) by nematicide usage in Michigan potato

 

 

 

 

 

production in 1988.
Chemical input Nematicide No Nematicide
Men 202.11 161.75
Phorphonrs 134.60 137.92
Potassium 221.82 208.00
Sulﬁrr 15.67 5.40

 

 

 

 

 

Table 29. Irrigation and rotation schedules in Michigan potato production by nematicide

 

 

 

 

 

 

 

 

usage in 1988.
Practice Nematicide No Nematicide
Irrigation
(portion of ﬁrms irrigated) 0.982 0.495
Rotation
ears out ofpotato) 1.750 1.250

Full Model

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q 500 ' r2 = 0.4400
3 rs
v 400 131
3 1:1 1:1
0 3m 0 T
3 ° 0 135 ’5' 5' Ia '9
8 20° ‘ W °
h- . O . El 0 ﬁ
'8 100 - ° 0 El Nematicide
‘3 . 9 No Nematicide
O I I I 7 ﬂ
E 0 100 200 300 400
Nitrogen (lb/A)
Figure 24. Correlation of nitrogen inputs to expected yields in Michigan
. potato production by nematicide usage in 1988
(Nematicide: y=314.56-0.02.x, r2=0.000; No Nematicide:
y=216.39-02x, r2=0.032).
Full Model
A 500 ' r2 = 0.4374
<1
3 no a
5 .
.9 300 - o
g 0 ” I t
‘3' 200 it'——_ o .
o. o 3 ..
'3 100 . o Nematrcrde
'8 9 9 9 No Nematicide
E. O I I ﬂ T I 1 A
0 1 2 3 4 5 6
Number of years out of potato
Figure 25 Correlation of rotation (years out of potato) to expected yields in

Michigan potato production by nematicide usage in 1988
(Nematicide: y=324.93-8.1x, r2=0.014; No Nematicide:
y=216.39-0.2, r2=0.032).

62

In genual, growers that used nematicides were also twice as likely to irrigate (Table
29). The correlation of portion of land irrigated to expected potato yields was positive for
both nematicide users and non-users alike, although nematicide users were much more
dependent on irrigation than non-users (Figure 26). Both correlations were statistically
signiﬁcant (nematicide: 13=0.283, p=0.016; no nematicide: r’=0.301, p=0.012). Correlating
both nematicide usage and portion of land irrigated simultaneously, increased the r1 to 0.597.

Sixty-ﬁve percent of the growers that did not use nematicides used at-plant
insecticides. Of the growers that used nematicides, 100% used at-plant insecticides, 75%
used at-plant nematicides, 20% used fumigants and 30% utilized chemigants (Table 30).
Growers that used nematicides, applied an average of 2.46 lb ai/A of Temik ISG; while the
non-users applied an average of 0.56 lb ai/A (Table 31). Correlating Temik 15G usage with
potato yields, nematicide users had a positive correlation with a low r2 (r2=0.018, p=0.576)
and no nematicide growers had a negative correlation with a higher r2 (r’=0. 164, p=0.077).
Correlating both nematicide usage and Temik ISG simultaneously with expected potato
yields, it was found that 47.85% of the variability was explained by both of these factors
(Figure 27).

Three multiple regression models were used in an eﬁ‘ort to better understand the
variability in potato yields. The ﬁrst model only incorporated pesticide inputs. The second
model included these inputs, in addition to the crop rotation schedule, irrigation, and nitrogen
inputs. The third model was a revised version of model two.

The chemical input model included a constant and four variables. This model

explained 44.99% ofthe variability in potato yields (n=40, «=35, k=5, s’=5189.8, r’=0.4499).

63

Full Model
r2 = 0.5972

  
 

0E

Nematicide

 

 

 

 

 

Expected potato yield (cwtlA)
§

 

 

 

9 No Nematicide
O I I I I I I I
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Portion of land irrigated
Figure 26 Correlation of portion of irrigated land to expected yields in
Michigan potato production by nematicide usage in 1988
Nematicide: y=-455.5+780.7x, r2=0.283; No Nematicide:
=147.7+87.9x, r2==0.301).
Full Model
5m . 1'2 = 0.4785
400 El I
300 o ,
B a Nematicide

9 No Nematicide

  

 

 

O I 1 If I —I

0 1 2 3
Temik 15G (lb/ailA)

Expected potato yield (cthA)
§

uh

Figure 27. Correlation of Temik 15G inputs to expected yields in Michigan
potato production by nematicide usage in 1988 (N ematicide:
y=284.09+10.9x, r2=0.018; Nematicide: y=207.18-28.5x, r2=0.164).

64

Table 30. Fraction of Michigan potato growers utilizing at-plant insecticides and
nematicides bygnematicide usage in 1988.

 

 

 

 

 

 

 

 

 

Chemical input Nematicide No Nematicide
At-plant insecticide 1.00 0.65
At-plant nematicide 0.75 0.00
Furnigant 0.20 0.00
Chemigant 0.30 0.00

 

Table 31. The utilization of at-plant nematicides for Michigan potato production by
nematicide usage in 1988.

 

 

 

 

 

Chemical input Nematicide No Nematicide
(ai/A) (ai/A)
Temik 156 2.46 0.56
Mocap 106 0.21 0.00
Furadan 0.15 0.02

 

 

 

 

65

No nematicide use: Equation 10a
Y = 232.12 - 4366*le
Nematicide use: Equation 10b
Y = 232.12 + 38.908"X2NF + 6.16"‘X2F + 119.06"'X2C
where:

X1 = No nematicide use
X2 = Nematicide

Y = Expected potato yield

API = 1 if at-plant insecticide applied, 0 otherwise

NF = 1 if non-fumigant applied, 0 otherwise

F = 1 if fumigant applied, 0 otherwise

C = 1 if chemigant applied, 0 otherwise
Of these ﬁve indicators, all except the fumigant were signiﬁcantly diﬁ‘erent ﬁom zero at the
p<0.20 level (Table 32). This model did not take into consideration anything other than
pesticide inputs. A larger model was developed that included years out of potato, portion of
land irrigated and nitrogen inputs as well as pesticide usage. This larger model explained
66.52% of the variability in expected yields (n=40, dﬁ29, k=11, s7=3812.2, rz=0.6652).
No nematicide use: Equation 118

Y = 129.19 + 12.00“'X1R + 103.15"'X1I - 725‘le

Nematicide use: Equation 11b

Y = 129.19 483*in + 260.61'X2, - 0.24"‘X2N

- 43.36‘X2Np -22.72*X2,. + 67.35"'X3C

where:

X1 = No nematicide use
X2 = Nematicide

66

Table 32. A multiple regression model correlating potato yields to pesticide input by
nematicide usage for Michigan potato growers.

 

 

 

 

 

 

 

 

 

 

 

Coeﬁicient Beta Estimator P-value
Constant B0 232.12 0.000
At plant insecticide 11 -43.66 0.164
Non-Funu'gant N2 38.91 0. 173

Lungigant F2 6.16 0.876
, Chemigagnt C2 119.06 0.001

 

67

Y = Expected potato yield

R = Number of years out of potato

I = Portion of land irrigated

API = 1 if at-plant insecticide applied, 0 otherwise

NF = 1 if non-ﬁrmigant applied, 0 otherwise

F = 1 if fumigant applied, 0 otherwise

C = 1 if chemigant applied, 0 otherwise
Of these eleven indicators, only four were statistically signiﬁcant at the p<0. 10 level (Table
33). This smaller model included irrigation and chemigation inputs along with a constant
(n=40, df--36, k=4, s’=3439.6, r2=0.6250).
No nematicide use: Equation 128

Y = 144.83 + 91.43"'X1I
Nematicide use: Equation 12b
Y = 144.83 + 145.25*X2, + 8158*ch

where:

X1 = No nematicide use
X2 = Nematicide

Y = Expected potato yield

I = Portion of land irrigated

C = 1 if chemigant applied, 0 otherwise
All four parameters were statistically signiﬁcant at the p<0.05 level (Table 34).

Of all the individual models presented, the strongest differences in economics is
between nematicide users and growers that do not use nematicides. Expected yield, expenses,
and expected proﬁt were signiﬁcantly diﬁ‘er‘ent ﬁom each other at the p<0.05 level (t-test,
Table 35). The expected yields of potato growers using nematicides averaged 310.75
cwt/acre while ﬁrmers that did not use nematicides reported yields that averaged 191.25

cwt/acre. Selling price was not statistically diﬁ'erent Gr=—6.09 for nematicide users and 'x=6.37

68

 

 

 

 

 

 

 

 

 

 

 

 

Table 33. A multiple regression model correlating potato yields to rotation schedule,
irrigation usage, and pesticide usage by nematicide use.
Coefﬁcient Beta Estimator P-value
Constant B0 129. 19 0.022
Rotation R1 12.00 0.768
”mule R2 -4.83 0.446
Irrigation 11 103.15 0.009
12 260.61 0.014
Nitrogen N1 0.00 0.360
inputs N2 024 0.996
At-plant APIl -7.25 0.806
insecticide
Non-ﬁrmigant NF2 -43.36 0.237
Mam F2 -22.72 0.551
, Chemigant C2 67.35 0.057

 

 

 

 

 

 

Table 34. A revised multiple regression model correlating potato yields to irrigation
usage and chemigation inputs by nematicide use.

 

 

 

 

 

 

 

 

 

 

Coeﬁcient Beta Estimator P-value
Constant B0 144.83 0.000
Irrigation 11 91.43 0.000

12 145.25 0.008
Chemigant C2 81.58 0.008

 

69

Table 35. The economics of Michigan potato production by nematicide use in 1988.

 

 

 

 

 

 

 

 

Economic indicator Nematicide No Nematicide
Expected yield“ 310.75 191.25
Expected sellirgprice 6.09 6.37
Expenses“ 649.37 466.49
Expected proﬁt“ 1217.41 722.20

 

‘Indicates a signiﬁcant diﬁ‘erence between nematicide and no nematicide use (t-test).

 

70

for non-users). Average expenses (x=649.37 for users and x=466.49 for non-users) and
aqrected proﬁt (i=1217.41 for users and 7F$72220 for non-users) were statistically different
(p<0.05 level). Their was a negative correlation when expenses (SIA) were regressed with
return over direct costs. When both expenses and nematicide use were correlated
simultaneously with return over direct cost, the model only explained 21.85% of the
variability in return over direct costs (Figure 28).

W Of the 40 growers surveyed 50% of them used a rotation scheme
with one year out of potato production (Table 36). This accounted for 63% of the surveyed
acreage. About 14% of the acreage was continuous potato, 18% two years out of potato,
and the remainder (ca 5% three or more years out). Four and ﬁve years out of potato were
not included in the linear model analysis because each had an n=1.

There were no signiﬁcant diﬁ‘erences among these rotation schemes on nutrient inputs
(Table 37). Nitrogen inputs averaged from 121.79 1b/A for growers that do not rotate to
193.30 for growers who are out of potato for a single year.

The portion of land irrigated ranged ﬁom 37.9% for growers that do not rotate, to
90% for growers out of potato for one year (Table 38). Correlating portion of land irrigated
to expected potato yields, each rotation scheme had a positive correlation (Figure 29). The
correlation was not statistically signiﬁcant for 0 years out of potato (r’=0.515, p=0. 172) and
3 years out of potato (r2=0.387, p=0. 141), but was statistically signiﬁcant for 1 year out of
potato (r’=0.539, p<0.001) and for two years out of potato (r'=0.568, p=0.03 8). Correlating
both portion of land irrigated and years out of potato with expected potato yields

simultaneously, the model explained 54.84% of the variability in potato yields.

71

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

Full Model
2 m a 3 r2 = 0.2185
5 .
a o s

g 2000 . E El Nematrcrde.
8 a 151 El No Nematicide
‘3'
.tr
'6
3'3
8
g .1” I I I I

200 400 600 800 1000

Expenses (SIA)

Figure 28. ' Correlation of direct costs associated with planting and
expected yields in Michigan potato production by
nematicide usage in 1988 (N ematicide: y=1632.2-0.64x, r2 =0.024;
No Nematicide: y=1377.8-1.41x, r2=0.150).

500 . Full Model

<2 , r2 = 0.5484

2 .m- z

E

O a

'5. 30° 1

2 ﬂ

3 2m Cont potato

° 1 t
m year on
g 100 D 2 years out
o 9 3 years out
5. 0 I I I I I r 1
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Portion of land irrigated

Figure 29. Correlation of portion of irrigated land to expected yields in
Michigan potato production by years out of potato in 1988
(Cont potato: y=127.46+117.4x, r2=0.515; 1 year out: y=29.334+
266.0x, r2=0.387; 2 years out: y=119.2+173.3x, r2=0.539; 3 years
out: y=237.5+112.5x, r2=0.568).

72

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 36. An estimate of total Michigan potato production by rotation schedule in 1988.
Rotation n Surveyed Percent Estimated
schedule acres surveyed MI production

(years out of (acres)
Dame)
0 5 1425 13.88 6038.14
1 20 6532 63.62 27677.96
2 8 1868 18.19 7915.25
3 5 444 4.32 1881.36
4 1 22 0.21 93.22
5 1 75 0.73 3 17.80
Table 37. The utilization of nutrient inputs (lb/A) by rotation schedule in Michigan
potato production in 1988.
Rotation Nitrogen Phosphonrs Potassium Sulfur
schedule
0 121.79 70.40 241.67 21.60
1 193.30 154.20 202.42 7.71
2 183.93 145.50 222.75 15.00
3 187.57 114.87 232.31 7.84
4 266.00 114.00 234.00 0.00
5 127.00 162.00 162.00 0.00

 

 

 

 

 

 

 

 

73

Table 38. Portion of ﬁrms under irrigation in Michigan potato production by rotation
schedule in 1988.

 

 

 

 

 

 

 

Rotation n Irrigation
schedule (portion of farms
irrigated)
0 5 0.379
1 20 0.900
2 8 0.704
3 5 0.600
4 1 0.000
5 1 1.000

 

 

 

 

 

7 4

In addition to irrigation, the impact of chemical nematicides and insecticides were
analyzed. At least 75% of the growers used an at-plant insecticide. Growers that did not
rotate, did not use nematicides. Growers that rotated out of potato at least one year used a
variety of types of nematicides (at-plant nematicides, fumigants and chenrigants). However,
a large portion of each category of grower did not use nematicides (Table 39). Temik 156
was most predominately used by growers that rotated out of potato for two years (Table 40).

A chemical scaling was developed to better understand the relationship between
pesticide usage and potato yields. Chemicals were assigned a numerical value based on the
toxicology of the pesticide. The following numerical values were assigned:

At-plant insecticide 1 or

At-plant nematicide 2

Fumigant 3

Chenrigant 4
The largest value a grower could receive was a 9, because growers were not assigned a
number for both an at-plant insecticide and at—plant nematicide. If enough of the pesticide
was applied at an appropriate level to have nematicidal properties a 2 was assigned (i.e., if
Temik 15 Gwas applied at the rate of 3.0 lb ai/A). The chemical scaling was correlated with
expected potato yields for the four diﬁ‘erent rotation schemes. It was found that all schemes
had a positive correlation with chemical scaling. The only regression that was statistically
signiﬁcant was for one year out of potato (r2=0.535, p<0.001). The other three regressions
(no rotation, two years out, and three years out) were not statistically signiﬁcant (r2=0.033,

p=0.771; r1=0.459, p=0.065; and r2=0.138, p=0.539, respectively).

 

 

 

 

 

 

 

8.; 83 82 n
8o... 82. 83 e
o8... 8.0 25 m
8o... 88 83 a
83 2:. 83 _
82. 888 888 o
28 33 23
5088 02 use: on. 888... 038:8 8833.

 

 

 

 

 

.82 s oases e882 2 88:88 888 50268 so 8.8882. sass 8 e888: 25 .8 see...

 

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e8.e 8% 8e 83 8.. m
83 88 8% 88 83 e
82. 83. 8S 8:. 83 n
38 one 83 as... 83 N
8:. 83 830 83 88 _
8._ 82. 88... .8... 82 e

88882 68882 oases

888:2 oz 33830 3.888 3.592 353-2 88.60.

 

.82 a oaeoes 88.2 3 88888 as 88885 age-.. 838. asses e88 58263 8 eases .8 6.89

76

 

 

 

 

 

A 500 Full Model
:5 . r2 = 0.4973
5 400 z ,
I; 3 o a II
1
'5. 30° _ o
e
"" El Cont otato
g 200 f 2’ o 1 P
a. , ll year out
'6 l 2 years out
100

g a 9 3 years out
g- 0 ﬁ I I I 1 I

0 2 4 6 8 10

Chemical scaling

Figure 30 Correlation of chemical nematicide usage and expected yields

in Michigan potato production by years out of potato in 1988
(Cont potato: y=150.0+27.5x, r2=0.033; 1 year out: y=195.3+34.6x,
r2=0.535; 2 years out: y=181.7+19.1x, r2=0.459; 3 years out:
268.2+20.5x, r2=0.138).

'77

A multiple regression model was utilized to assess the impact of irrigation and

chemical scaling by rotation schedule. The following equation has twelve estimated values

(n=38, df=27, k=11, s’=3317.6, r’=0.702):
For continuous potato: Equation 138
Y = 127.46 + 117.44“X0l
For one year out of potato: Equation 13b
Y = 44.98 + 194.86"'X1l + 23.48‘Xls
For two years out of potato: Equation 130
Y = 126.64 + 120.07'X2, + 9.62"‘X2s
For three years out of potato: Equation 13d
Y = 237.50 + 175"‘X3I - 31.25"'X3s
where:
X0 = Continuous potato
X1 = One year out of potato
X2 = Two years out of potato
X3 = Three years out of potato
Y = Expected potato yield
I = Portion of land irrigated

S = Numerical value between 0 and 9, depending on
pesticide applications

Of the eleven parameters tested, nine of them were signiﬁcant at the p<0.20 level.

The only two parameters that did not pass this evaluation were the constant for one year out

of potatoes, and the chemical scaling for two years out of potatoes (Table 41).

Exdudingﬂresetwopamnetasﬁomﬁemddnwasfoundﬂrataﬂparameterswere

at least signiﬁcant at the p<0.20 level (Table 42). The equations for this model are as follows

(n=38, df=29, k=9, s2=3284.2, r==0.753).

78

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 41. A multiple regression model correlating potato yields to irrigation and
pesticide scaling by rotation schedule.
Coeﬁicient Beta Estimator P-value
B0 127 .46 0.002
Constant B1 44.98 0.460
B2 126.64 0.009
B3 237.50 0.000
10 1 17.44 0.102
Irrigation 11 194.86 0.009
12 120.07 0.123
13 175.00 0.014
C1 23.48 0.003
meal C2 9.62 0.294
C3 -31.25 0.137
Table 42. A revised multiple regression model correlating potato yields to irrigation and
pesticide scaling by rotation schedule.

Coeﬁcient Beta Estimator P-value

B0 127 .46 0.002

Constant B2 119.18 0.012
B3 237 .50 0.000

10 1 17 .44 0. 100

Irrigation 11 243.55 0.000
12 173.34 0.005

13 175.00 0.013

Chemical C1 23.08 0.003
”ding C3 -3125 0.134

 

 

 

 

 

 

 

 

79

For continuous potato: Equation 148
Y = 127.46 + 117.44"‘X0I
For one year out of potato: Equation 14b
Y = 243.55"'X1I + 23.04"'X13
For two years out of potato: Equation 140
Y = 199.18 + 173.34"'X2I
For three years out of potato: Equation 14d

Y = 237.50 + 175*X3, - 31.25'X33

X0 = Continuous potato

X1 = One year out of potato
X2 = Two years out of potato
X3 = Three years out ofpotato

Y = Expected potato yield

1 = Portion of land irrigated

S = Numerical value between 0 and 9, depending on

pesticide applications

When reviewing the economics of potato production by rotation scheme, mean potato
yields ranged ﬁom 80 chA in the one ﬁrm that was four years out of potato to 305 chA
for growers that were three years out of potato (n=5). Selling prices ranged from a low of
$5.23 to $11.00 per cwt. Expected proﬁt ranged ﬁom $73.17 to $1107.93 (Table 43).

When correlating the direct costs associated with potato production with returns over
direct cost, continuous potato and one year out of potato had a positive correlation.

However, expenses did not explain very much ofthe variability in return over direct costs and

these correlations were not statistically signiﬁcant at the p<0.05 level (For continuous potato:

80

 

 

 

 

 

 

 

Table 43. The economics of Michigan potato production by rotation schedule in 1988.
Expected Expected Expected
Region yield selling price Expenses proﬁt
(cwt/A) (Slcwt) (SIA) ($/A)
0 172.00 5.23 444.52 457.28
1 269.75 6.05 544.75 1107.93
2 241.25 6.81 531.96 976.38
3 205.00 6.10 720.13 1099.87
4 80.00 11.00 806.83 73.17
5 250.00 6.00 536.43 963.57

 

 

 

 

 

 

 

81

r2=0.036, p=0.759); and For one year out of potato: r7=0.122; p=0.132). The correlation was
negative for two and three years out of potato. The negative correlation was not signiﬁcant
for two years out of potato (1"=0.043, p=0.621). The negative correlation was signiﬁcant for
three years out of potato and this model explained much more of the variability in return over

direct costs (r1=0.798, p=0.041, Figure 31).

DISCUSSION

In comparing the results of the models, it is necessary to divide the models into two
basic types. First the series of single effect models will be examined, then a series of multiple
regression models.

WWW. Comparing the r2 values for the
various models that were developed, the strongest relationship between expected yields and
a single-eﬁ‘ect is when the data are stratiﬁed into regions of the state of Michigan (Table 44).
The 13's range ﬁom 0.462 for Tenrik to 0.836 for chemigation usage as the single eﬁ‘ect.
Hence, if one had to estimate potato yields with only one input, proportion of chemigation
usage would be the input of choice if that ﬁrm was located in the northern or north central
portion of the state. In all other parts of the state, irrigation would be the best indicator.
Chanigation estimates are limited to the north and west central regions because those were
the only regions reporting chemigation use in the survey. However, the southwest portion
of the state was not included in this analysis because there were only two ﬁnners surveyed
in this area and they both reported the same yield. Hence, for growers in the south western
part of the state the best indicator would be portion of land irrigated if their nematide usage

was known or irrigation based on ﬁrm size, if that parameter was known. These r’ were

O
a

 

82

 

 

 

 

Return over direct costs ($IA)
§ §

° Full Model
0 . r2 = 0.3556
0
ﬂ 0
0
A A.
0,, e , V‘" 0 Cont potato
.0 .9 .9 9 1 year out
El ﬂ 2 years out
9 3 years out
200 400 600 800 1000
Expenses (SIA)

'11
ﬂ.
OD
l—l

Correlation of direct costs associated with planting and

expected yields in Michigan potato production by years

out of potato in 1988 (Cont potato: y=27.8+1.0x, r2 =0.036;
1 year out: y=306.6+1.5x, r2=0.122; 2 years out: y=1213.4+
0.4x, r2 =0.043; 3 years out: y=3565.6-3.4x, r2=0.798).

83

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 44. A summary of the single eﬁ‘ect models for potato yield prediction for the state
as a whole, or by ﬁrm size, region, nematicide use, or rotation schedule (r’).
Eﬁ'ect MI Farm size Region Nema- Rotation
ticide
Irrigation 0.4370 0.5572 0.7191 0.5972 0.5484
Rotation 0.0070 0.2865 0.5307 0.4374 --
Temik 0.1720 0.3631 0.4621 -- -
Cherrrigantl 0.3110 0.3318 0.8357 - --
getting 0.0000 .- -- -- 0.4973
Expenses regressed with return over direct cost (proﬁt)
Ewes J 0.0000 0.2501 0.3102 0.2185 0.3556

 

 

 

84

generated on the basis of the best overall indicator considering a group of ﬁrms, for example
this might be a good statistic for a state cooperative extension agent to use if the goal was to
assess the potential yield for 20 growers.

Ifa single grower was to use these models, and would like to assess the advantage of
irrigation, chemigation, or rotation, then that grower would more than likely use the single
regression lines presented in the ﬁgures. For example, if a medium size grower in Montcalm
county wanted to assess the impact of irrigation on the ﬁrm, one would examine the models
for region and for size that regressed portion of land irrigated with expected yield and use the
model with the highest r2 value. Regressing portion of land irrigated with expected yield by
ﬁrm size, one ﬁnds an r2 of 0.320 for medium sized ﬁlms (Figure 14) and 0.119 for ﬁrms
in Montcalm County (Region 3, Figure 20). Therefore, it would be better to use the model
associated with ﬁrm size. And estimate that going ﬁ'om O to 100% of the ﬁrm irrigated
would roughly double yield (y=165.1 + 144.96*x, Figure 14).

In addition, it is interesting to look at the impact of stratifying the data on models
regressing return over direct costs with expenses. In the state as a whole, there does not
appear to be a relationship between the two. However, after stratifying the data into ﬁrm
size, region, nematicide use and rotation schedule, it is found that the strongest relationship
between expenses and expected proﬁts (excluding ﬁxed costs) is when the data are stratiﬁed
along the lines of years out of potato (Table 44).

MW. Similar to the single eﬂ‘ect models or simple linear
models, region models provided the best indicator of yield. The large region model gave the
best r2 value, which is not too surprising because it also had the largest number ofparameters

in the model (Table 45). However, the version with fewer parameters cut the number of

Table 45.

85

A summary of the multiple regression models for the state of MI, and by ﬁrm
size, region, nematicide usage, and rotation schedule.

 

Model

r2

ﬁ

No. of parameters

Inputs required

 

MI Model

0.639

9

N-P-K-S
Irrigation
Rotation
Temik
Chemigation

 

Farmsize

0.689

12

Rotation
Irrigation
Temik
Chemigation

 

Smallerﬁrm size

0.647

Rotation
Irrigation
Chemigation

 

Region

0.815

.16

Rotation
Irrigation
Temik
Chemigation

 

Smaller region

0.799

Rotation
Irrigation
Chemigation

 

Nernaticides
alone

0.450

API
Nonfumigant
Fumigant
Chemiggtion

 

Nematicide

0.665

11

Rotation
Irrigation
Nitrogen
API
Nonﬁmrigant
Fumigant
Chemigation

 

Smaller nematicide

0.625

Irrigation
Chemigation

 

Rotation

0.702

11

Irrigation
Chemical scaling .

 

Smaller rotation

 

 

0.753

 

 

Irrigation
Chemical scaling

 

 

86

parameters in more than half, and still provided a reasonably accurate r1. Once again, the
region models did not apply to the southeast portion of the state, here the best model would
be the smaller version of the rotation schedule model (r2 = 0.753).

Sum. In practice, each of these models provided a ﬁirly good r2 value and an
accompanying p-value < 0.001. Therefore, when assessing the most appropriate model to
use it is important to examine both the inputs, and the ease, reliability, and cost of obtaining
that information, as well as the r2 value. For example, the difference in r2 between the small
nematicide model and the large region model is about 0.20; however, the number of separate
parameters is reduced ﬁom 16 to 4. Hence, it may be worth it to sacriﬁce some accuracy for

the economics of collecting the data or for emcacy of data collection.

ASSESSMENT OF INTRA- AND INTER- W W CULTIVAR

RESPONSES TO mm W
AND YERIICILLIIIM DAHUAE

INTRODUCTION

The root-lesion nematode, Manchu: menus (Cobb, 1917) Filipjev and
Schurrmans Stekhoven, 1941, and the fungus, micmimn dahliae, cause a disease
complex of potato (Rowe et al. 1988, MacGuidwin and Rouse 1990) which usually results
in low in tuber initiation, growth and development. Therefore, the combined presence of
these organisms make them an important economic pest complex in the north central
region of the United States (Kotcon et al. 1984, Rowe 1984). The disease symptoms
often go undetected because the potato plant appears healthy until very late in the growing
season. Infected potato plants usually senwe about two weeks earlier than normal plants,
resulting in low potato yields. In addition, improved cultivars and management techniques
can often mask the yield loss (Rowe et al. 1987).

This disease complex can have a synergistic effect on potato tuber production
(Riedel and Rowe 1985). Although the exact mechanisms of the interaction are unknown,
several other Wench“: species (e.g., B. scrihneri and R. mama have been tested, but
none have produced the synergism associated with 2. mm (Riedel and Rowe 1985).
Hence, it is believed that the synergistic reaction is not due to wounding.

87

88

The amount of yield loss has been very diﬁicult to predict (Francl et al. 1987,
Johnson 1988, Wheeler et al. 1994), and oﬁen times varies signiﬁcantly ﬁ'om year to year
(Rowe et al. 1985).

One possible management approach to the control of early die is the use of
resistant cultivars (Davis 1985). The following research was an eﬂ‘ort to identify the
synergistic relationship between the potato early die causal agents and ten diﬁ‘erent potato

cultivars in Michigan and quantitatively compare the relationship among cultivars.

OBJECTIVES
The objective of this study was to assess the potential damage ﬁom potato early

die both within each cultivar and between cultivars and to assess population densities of
Enrolment: penetrans and Winn) dahliae.

METHODOLOGY

Three experiments were initiated in an attempt to quantify the response of ten
cultivars to predetermined inoculated levels of mm mm (PP) and
Menigillinm dahliae (VD). The ﬁrst experimentutilized microtiles, the second was a
greenhouse study, and the third was a ﬁeld study.
Inoculum preparation

mm: mm was cultured at Michigan State University and was
originally obtained in Wisconsin. It was cultured on alﬁlﬁ callous in tubes.

The Wm dahliae was also received from \Vrsconsin. Plugs of y. dahliae

grown on alcohol-strep medium were transferred to 250 ml ﬂasks containing 50 ml of

89

Czapek's Box Broth media. Flasks were plugged, placed on a shaker for four days. Then
1-2 ml aliquots were transferred to 100x15 ml petri plates containing 0.25 strength potato
dexdrose agar. Plates were placed in dark at room temperature for 14 to 21 days before
inoculating potato tubers.
Jolly Road Microtile Study

Ten cultivars of potato were inoculated with four treatments (check, PP, VD,
PP/VD) in a randomized block design replicated nine times at a research site on the
campus of Michigan State University. A sandy loam soil was used. It was treated with
methyl bromide prior to the study and was placed in the microtiles prior to planting. The
tubers were placed in ten liter microtiles with one non-treated tuber planted in between
each of the microtiles. Nematodes were transferred within the alfalﬁ callous and placed
directly on the tuber. Verticillium was scraped ﬁ'om the petri plate directly onto the tuber
as well. The control treatment had pathogen-free callous and medium placed on the tuber.
Herbicides and insecticides were used throughout the growing season as needed. The
tubers were inoculated with 39 nematodes/ 100 cc of soil and 700 propagules of
Verticillium/100 cc of soil.
MSU Greenhouse Study

Five of the previous ten potato cultivars were inoculated with ﬁve treatments
(control, low PP, high PP, VD, PP/VD) in a randomized block design replicated thirteen
times. Sandy loam soil was steam sterilized prior to the inoculation. Both the nematodes
and Verticillium were put into slurries and aliquots were injected into the soil. The pots

contained 2500 cc of soil. The low nematode treatment contained 28 PP per 100 cc of

9O

soil, with the high level averaging 280 nematodes. Verticillium levels were 2534
propagules per 100 cc of soil.
Montcalm County Potato Research Farm Field Study

The other ﬁve cultivars were used for this study (Red Dale, Kennebec, Hudson,
Desire and Rosa). A randomized block design was replaced thirteen times. Soil was
fumigated with methyl bromide before planting. Each tuber was placed directly in the soil
with a buﬁ‘er tuber planted in between each experimental unit to act as a barrier. No
artiﬁcial barrier was placed between tubers. It was estimated that each plant had 20 liters
of soil in direct association. Therefore, estimated inoculum levels were 3.5 PP per 100 cc
of soil at the low treatment level, 35 PP at the high level, and 316 propagules of
Verticillium per 100 cc of soil.

Statistical analysis

Yuber yields (g/plant), the number of Manchu: ms (1.0 g of root tissue
and 100 cc of soil) and propagules of modular) dabliae (1 g of stem tissue) were
quantiﬁed at harvest. Soil nematode counts were processed through centriﬁrgal-ﬂotation
technique (Jenkins 1964). Potato root samples were cut, emerged in a mercury solution
and placed on a shaker for 48 hr (Bird 1971).

The data were analyzed utilizing two different methods, depending on the
objective of the analysis. When the objective was intra—cultivar speciﬁc, regression
analysis was utilized in an effort to see if the nematode and Verticillium had individual or
joint impact on tuber yields. A series of four different regression models were generated
for each site. The ﬁrst two models used the data points for PP and VD as they were

quantiﬁed. The ﬁrst model was a multiple regression model that had four parameters:

91

constant, PP, VD, and PP*VD. The second model had two parameters: constant and
PP‘VD.

The other two models used a transformation of the PP and VD datum points. The
natural log (n+1) was used because it adhered to the hypothesis that the more of either
pathogen that is present, the less impact each individual pathogen will have on the yield
(Francl et al. 1987). The models were slightly diﬁ‘erent at Montcalm, because there were
no root nematode data. By the time of harvest, the roots had deteriorated on some of the
cultivars. Initially, the data were analyzed using the soil PP counts and VD, but the
interaction terms were rejected so strongly that in all the cultivars (with p-values ranging
from 0.406 to 0.826, Table 13’) that the models were analyzed without that term
altogether. The second objective was inter-cultivar speciﬁc, and a series of ANOVAS
were utilized to assess the relative impact of the pathogens on the ten diﬁ‘erent cultivars

studied. SYSTAT and Minitab were the two statistical packages utilized in this study.

RESULTS

The ﬁrst question that needed to be addressed was whether the inoculations were
successful. A three-way ANOVA was used, dividing the source of variation among
cultivar, treatment and replication for P. penetms and M. dahliae fora harvest sampling
date. At all three sites nematode inoculation was ﬁirly successful as evidenced by the low
p—value for treatment (<0.001, Table 1) and the means themselves (Table 2). The plots

were contaminated with Verticillium by harvest (Table 3) at all three sites. The

 

1

All tables are numbered in relation to the sequence of experiments performed. They are
found at the end of the chapter.

92

contamination was greatest at Jolly Road and least in the greenhouse (Table 1). Since
there was a great deal of contamination, a series of regressions were used, along with
standard mean separation techniques.
Iota-cultivar specific observations

Redﬂale. The cultivar Red Dale was evaluated at Jolly Road and at Montcalm.
Yield averages ranged ﬁ'om 259 g/plant in the PPND treatment to 321 g/plant in the
check (Table 4). A two-way ANOVA with treatment and replication as sources of
variation returned a p-value of 0.587 (Table 5), which means there is a lot of variability in
the replications and one could accept the null hypothesis (there are no differences among
treatments) and be correct about 58.7% of the time. At Montcalm, in the ﬁeld study, all
of the treatments out-performed the nematode and verticillium-free control (Table 4).
Although the likelihood of rejecting the null hypothesis was much larger (p-valuFO.124,
Table 5), the yield diﬁ‘erences were atypical. In addition, the data were used to ﬁt
multiple regression models for each cultivar. The PP values ranged ﬁ'om 0 to 406.9 PP/g
of root tissue and 100 cc of soil at Jolly Road (Table 6) and ﬁom 0 to 14 PP/100 cc of soil
at Montcalm (Table 8). Verticillium ranged from 0 to 3383.3 propaguleng stem tissue
(ppg) at Jolly Road (Table 6) and ﬁom 0 to 15 ppg stem tissue at Montcalm (Table 8).
The best ﬁt regression line at the Jolly Road site came ﬁ'om the natural log transformations
which indicated that there was enough population present to cause less damage per
individual pathogen (Tables 9 and 10). The best ﬁt had an r2 of 0. 179 associated with it
(Table 10). Likewise, with the interaction term only, the natural log had a better ﬁt
(r’=0.069). The regression model has a negative slope (decrease in yield) for PP and a

positive slope for VD and VD’PP. Although the r2 was less at Montcalm, the slope trend

93

was similar in that the only negative parameter was PP in any of the models (Tables 13
and 14).

Kennebec. The cultivar Kennebec was evaluated at Jolly Road and at Montcalm.
Yield averages ranged ﬁom 261 g/plant in the PPND to 318 g/plant in the PP only plot at
Jolly Road (Table 4). A two-way AN OVA with treatment and replication as sources of
variation returned a p-value of 0.754, which meant that there were no differences among
treatments (Table 5). At Montcalm, the yield averages ranged ﬁom 1886 in the PPND
treatment to 2401 g/plant in the check (Table 4). Here, the ANOVA returned a p-value of
0.495 (Table 5). Although variance was too high to give statistical signiﬁcance to the
interpretation, in both trials the lowest yield averages were in the PP/VD plot.

The range of values for the linear models ﬁ'om the Jolly Road data was from 0 to
636 PP/100 cc of soil and 1 g of root tissue, 33 to 4367 VD ppg of stem tissue, and the
joint interaction term ranged ﬁom 0 to ca. 2 million (Table 6). At the Montcalm site, the
PP values ranged ﬁom 0 to 7 PP/ 100 cc of soil and VD ranged ﬁom 0 to 52.2 ppg of stem
tissue (Table 8). The best ﬁt regression model for Jolly Road resulted in an r2 of0. 156
(Table 9). It used non-transformed data. The only negative slope was ﬁom the PP'VD
interaction parameter. However, the parameter was extremely close to zero. At
Montcalm, the best ﬁt r2 was much lower (r’=0.032, Table 14). It was obtained ﬁ'om the
transformed data. Both the PP and VD parameters had negative estimators. .

Sunder. The cultivar Superior was evaluated at Jolly Road and in the
greenhouse. Yield averages ranged ﬁom 238 g/plant in the PPND treatment to 296

g/plant in the VD only plots at Jolly Road. In the greenhouse, yields were much lower

than at either ofthe other two sites, they ranged from 11.6 in the VD only plot to 23.1

94

g/plant in the check. An AN OVA detected no treatment differences at Jolly Road (p.
value=0.897), but there may have been diﬁ‘erences in the greenhouse (p-value=0.102).

To ﬁt the regression models, PP ranged ﬁ'om 0 to 900 nematodes per 1 g of root
tissue and 100 cc of soil and VD ranged ﬁom 0 to 2850 ppg of stem tissue. Their
interaction ranged ﬁ'om 0 to 1.9 million (Table 6). In the greenhouse, PP ranged from 0.0
to 560 while VD ranged ﬁ'om 0 to 39.4 propagules. The interaction term ranged ﬁ'om 0.0
to 4761 (Table 7). The best ﬁt model for Jolly Road was with the untransformed data
(r2=0. 194, Table 9). In this case, the only parameter that was negative was the interaction
term. Although the interaction parameter coeﬁicient was small (-0.001), it had a p—value
of 0.061 associated with it, which meant that it was nearly statistically signiﬁcant (p~
value=0.05). The ﬁt was much lower for the greenhouse regression (r’=0.05 1), but the
untransformed data ﬁt better than the natural log transformation as well. However, none
of the pathogen parameters had negative coeﬁcients in the natural log transfer.

W. The cultivar Russet Burbank was evaluated at Jolly Road and at
the greenhouse. Yield averages ranged ﬁ'om 210 g/plant in the PPND plots to 353
g/plant for the check treatment at Jolly Road (Table 4). The two way ANOVA
partitioning the variance into treatment, replication and error resulted in a signiﬁcant p-
value of 0.037 (Table 5). In the greenhouse, yield averages ranged ﬁom 0.5 in the PPND
treatment pots to 4.2 g/plant in the low PP plot (Table 4). The ANOVA was much less
likely to reject the null hypothesis for the greenhouse study (p=0.549, Table 5).

The data set for the multiple linear regression models for the Jolly Road site
ranged ﬁom 0 to 550 PP/l g ofroot tissue and 100 cc ofsoil with VD ranging from 16.7

to 2750 propagules per gram of stem tissue. Their joint interaction ranged ﬁom 0 to

95

273,000 (Table 6). Although both ﬁll models had a negative coefﬁcient of joint
interaction, the transformed data provided a slightly better ﬁt (Tables 9 and 10). In the
greenhouse, the range for PP was ﬁ'om O to 823 nematodes per gram of root tissue plus
100 cc of soil, while VD went ﬁ'om 0 to 19 propagules per g of stem tissue. Their joint
interaction ranged from 0 to 368 (Table 7). Neither regression ﬁt as well as the previous
data set (r2=0.036 for untransformed data; r1=0.075 for transformed data set, Tables 11
and 12).

W. The cultivar Norkota Russett was evaluated as Jolly Road and in
the greenhouse. Yield averages at Jolly Road ranged ﬁom 270 g/plant in the VD only plot
to 314 yplant in the check. There were no differences among treatments (p-value=0.774,
Table 5). In the greenhouse, the ranges were ﬁom 16.3 in the PP only to 24.0 g/plant in
the check. The probability of Type I error (rejecting the null hypothesis, when it should
have been accepted), was much lower in this experiment (p-value=0.324).

At harvest PP ranged from 0 to 171 and VD from 0 to 1668 at the Jolly Road site,
while their joint interaction ranged from 0 to ca. 200,000 (Table 6). The best ﬁt was with
the transformed data (r2=0.117, Table 10). In this case, both pathogens had negative
coefﬁcients associated with their parameters and the interaction had a positive coeﬁcient.
In the greenhouse, the pathogen counts ranged ﬁ'om 0 to 288 PP/l g of root tissue and
100 cc of soil and 0 to 18.6 VD/l g of stem tissue, with the joint interaction ranging ﬁom
0 to 297.6. The best ﬁt linear model once again was with transformed data. However,
this time the only negative coeﬁcient was with the interaction parameter. Both others

were positive. The ﬁt was not as good (pevalue=0.070, Table 12).

96

Hudson. The cultivar Hudson was evaluated at Jolly Road and at Montcalm. At
Jolly Road the average yields ranged from 212 g/plant in the PPND plots to 350 g/plant
in the VD only treatment (p=0.200, Table 5). Similar results occurred at Montcalm, the
lowest yield was also in the PP/VD with 1900 and the highest in the VD only (1900
yplant). However, this experiment had more variability associated with it (p=0.710,
Table 5).

At Jolly Road, at harvest PP counts ranged ﬁom 0 to 636 per 1 g of root tissue
plus 100 cc of soil, while VD counts ranged from 0 to 7050 propagules/1 g of stem, which
was almost twice the level in any other cultivar (Table 6). The natural log transformed
data provided the best ﬁt (r"=0. 189, Table 10), with negative coeﬁcient parameters for PP
and VD, and a positive coeﬁcient for the interaction. The PP range at Montcalm was
from 0 to 8.0 nematodes per 1 g of root tissue and 100 cc of soil, while the VD ranged
from 0 to 29.8 propagules/1 g of stem tissue. The best ﬁt at Montcalm was with the
untransformed data (r2=0. 125, Table 14). The only negative coeﬁcient was with the PP
parameter.

Desiree. The cultivar Desiree was evaluated at Jolly Road and at Montcalm. At
Jolly Road, the average tuber yields ranged ﬁ'om 230 g/plant in the PP only plot to 301
g/plant in the VD only treatment (p-value=0.624, Table 5). At Montcalm, the lowest
average yield was in the check and the highest was also in the VD only (Table 4). The p-
value was lower than at Jolly Road (p-value=0.292, Table 5).

PP ranged ﬁom 0 to 171 per g ofroot tissue and 100 cc ofsoil at Jolly Road with
VD ranging ﬁ'om 33 to 4100 propagules/g of stem tissue. Their joint interaction ranged

ﬁom 0 to ca 2.5 million. The best ﬁt linear model for Desiree was with the natural log

9 7
transformed data In ﬁct, this model explained more of the variability in tuber yield than
any of the other models for any of the cultivars (r2=0.347). The model was statistically
signiﬁcant. Both PP and VD had negative coeﬁcients for their parameters, with a positive
interaction coeﬁcient. At Montcalm, PP ranged ﬁ'om 0 to 5.0/ 100 cc of soil. VD ranged
from 0 to 32.4 propagules/g of stem tissue. Joint interaction ranged from 0 to 162. None
of the models ﬁt the data,.all three models had p-values greater than 0.975 (Tables 13 and
14).

Rosa. The cultivar Rosa was evaluated at Jolly Road and at Montcalm. Tuber
yield averages at Jolly Road ranged ﬁom 284 g/plant in the VD only plot to 354 g/plant in
the check (p-value=0.844, Table 5). At Montcalm, the average yield was the lowest in the
PPND plot (513 g/plant) and the highest in the low PP (1117 g/plant). The p-value was
much lower for Montcalm (p-value=0. 144).

The PP ranged ﬁom 0 to 500/g of root tissue and 100 cc of soil at Jolly Road with
VD ranging ﬁom 0 to 3250 propagules/g of stem tissue. Joint interaction ranged from 0
to ca 600,000.(T able 6). The best ﬁt regression model was with untransformed data
(r’=0.118, Table 9). However, the only negative coeﬁcient was with the interaction term,
and it was close to 0.0, with a high p—value associated with it (Table 9). At Montcalm,
PP ranged from 0 to 11/ 100 cc of soil and VD ranged from 0 to 37 propagules/g stem
tissue. Their joint interaction ranged from 0 to 68 (Table 8). The r2 was much lower than
at Jolly Road (r’=0.065) and was best with the untransformed data (Table 14). At
Montcalm, both PP and VD had negative coeﬁcient, and there was no interaction term.

Snowden. The cultivar Snowden was evaluated at Jolly Road and in the

greenhouse. Tuber average yields at Jolly Road ranged from 287 g/plant in the PPND

98

plot to 394 g/plant in the check (p-value=0.557, Table 5). In the greenhouse, the lowest
yield was in the VD only (14.7 g/plant), and the highest in the PPND (27.0 g/plant, Table
4). Although these diﬁ‘erences were atypical, the p-value was much lower than at Jolly
Road (p-value=0.314, Table 5).

The ranges for the linear regression model were 0 to 537 PP/l g of root tissue and
100 cc of soil and 0 to 2467 propagules of VD/g of stem tissue. The joint interaction
ranged ﬁ'om 0 to ca. 600,000. The natural log transformation provided the best ﬁt, with
an r’—=0.254 (Table 10). The coefﬁcient for the PP parameter had the only negative
coeﬁcient. The greenhouse ranges were 0 to 347 PP/ 1 g of root tissue and 100 cc of soil
and 0 to 46.4 propagules of VD/g of stem tissue. Their interaction ranged ﬁ'om 0 to ca
14,000. This was the broadest range in the greenhouse study. Similarly to the Jolly
Road study, the best ﬁt was the natural log transformation (r2=0.117, Table 12).
Likewise, the only negative parameter was the coeﬁicient for PP.

Atlantic. The cultivar Atlantic was evaluated at Jolly Road and in the greenhouse.
Yield averages at Jolly Road ranged ﬁom 179 g/plant in the VD only to 268 g/plant in the
check (p-value=0.471, Table 5). In the greenhouse, the. lowest yields were in the low PP
treatment, and the highest yields were in the PPND plot (Table 4). Treatment differences
in the greenhouse were more pronounced than at the Jolly Road site (p-value=0.23 9,
Table 5). .

At the Jolly Road site, the PP ranged ﬁ'om 0 to 500/g of root tissue and 100 cc of
soil. VD ranged from 0 to 4000 propagules/g of stem tissue. Their joint interaction
ranged ﬁ'om 0 to ca. 300,000 (Table 6). The best ﬁt regression model resulted with the

natural log transformed data (r2=0. 127). Both PP and the interaction coeﬁcients had

99

negative values. However, neither value had a very strong p-value associated with it
(Table 10). PP ranged ﬁ'om 0 to 502/g of root tissue and 100 cc of soil in the greenhouse,
while VD went ﬁom 0 to 23.4 propagules/g stem tissue. Their joint interaction ranged
from 0 to 220. The best ﬁt resulted from the untransformed linear model (#0066, Table
11). With this model, only the VD coeﬁcient was negative.

Inter-cultivar speciﬁc ﬁndings

There are several methods that can be used to compare across cultivars. Relative
yield (Table 15), relative ranking of tuber yields (Table 16) and percent yield loss (both
total yield and marketable yield, Tables 17, 18 and 19) were used to provide inter- and
intra-cultivar comparisons. In order to test pairwise signiﬁcance, number of replications
have to be utilized (Table 21). Relative yield was obtained by dividing each yield by the
largest yield on a cultivar by cultivar basis. The potential range, therefore, could go from
0 to 1, with 1 being the maximum by deﬁnition, and zero being the minimum only if a plot
contained no yield. Ranking was done by ordering the yields from lowest to highest, and
numbering them from 1 to n. Yield loss was done on a replication by replication basis.
The treatment was subtracted ﬁom the control. Therefore, a negative yield 1088 would be
in cases where the treated plot did better than the check.

W. The variance in relative yields was assessed utilizing a three way
ANOVA with cultivar, treatment, and replication as sources of variation. All three
sources had statistically signiﬁcant p-values associated with their f-statistics. The same
was true for the greenhouse. At Montcalm, the treatment source of variation had a larger
p-value (p-value==0.254, Table 1). The MSE ﬁom the ANOVA was used in pairwise

comparisons using the t-test to determine statistical signiﬁcance (p-value <= 0.05).

100

Jolly Road site. Russet Burbank is the only cultivar that had statistically
signiﬁcant (p-value <= 0.05) intra-speciﬁc differences among treatments. All three
treatments diﬁ‘ered ﬁom the check. Examining inter-cultivar diﬁ‘erences among the
diﬁ‘erent treatments, it can be seen that the greatest number of pairwise differences can be
found in the check. Both Russet Burbank (0.669) and Atlantic (0.663) are statistically
diﬁ‘erent ﬁom Superior (0.414), Hudson (0.385) and Rosa (0.369). Kennebec (0.626)
was diﬁ‘erent from Hudson and R088. Russet norkota (0.594) diﬁ‘ered only from Rosa. In
the low PP treatment, Kennebec (0.652) differed ﬁom Rosa (0.342), Hudson (0.359),
Superior (0.399), Snowden (0.406), and Russet Burbank (0.409). In the VD treatment,
only two cultivars diﬁ‘ered ﬁom each other, Desiree (0.569) and Rosa (0.297). In the
interaction plat (PPND), Norkota Russett (0.567) differed ﬁom Rosa (0.314) and
Hudson (0.324). Kennebec (0.536) also diﬁ‘ered ﬁom Rosa (Table 15).

Greenhouse study. The only statistical intra-cultivar speciﬁc diﬁ‘erence
occurred in Atlantic, where the check (0.420) and PPND (0.427) differed ﬁom the low
PP (0.230). The cultivar Russet Burbank diﬁ‘ered ﬁom the other four cultivars in all ﬁve
of the treatments (Table 15). There were no inter-cultivar speciﬁc signiﬁcant pairwise
diﬂ‘erences among the other four cultivars in the check, low PP, high PP, or PPND
treatments. In the VD only treatment, Snowden (0.426) and Atlantic (0.380) diﬁ‘ered
ﬁom Superior (0.199). ‘

Montcalm research site. There were two statistically signiﬁcant intra-cultivar
speciﬁc diﬂ‘erences at this research site. The VD treatment in the cv. Desiree (0.542) was
diﬁ’erent from the check (0.344); and the low PP in the cv. Rosa (0.373) was diﬁ‘erent

from the PP/VD treatment (0.171). There were no inter-cultivar differences in the check

101

and low PP treatments. In the high PP treatment, Desiree (0.461) was diﬁ‘erent ﬁom Rosa
(0.249). The relative yields of Red Dale (0.568) and Desiree (0.542) were greater than
Rosa (0.232) in the VD only treatment. In the PPND treatment, Desiree (0.498), Red
Dale (0.453) and Kennebec (0.358) had statistically diﬁ‘erent relative yields from Rosa

(0. 17 1).

Welds. A three way analysis of variance, distributing the
variance among cultivar, treatment and replication, was run on the ranked values for tuber
yields. This resulted in no signiﬁcant statistical differences among cultivars. Treatments
had lower p-values than cultivar diﬁ'erences, but not as low as the p-values for replication.
The p-values for replication were statistically signiﬁcant at all three locations (Table 1).

Jolly Road site. At the Jolly Road site the only cultivar that had statistically
signiﬁcant intra-cultivar diﬂ‘erences was Russet Burbank. Similariy to relative yield, all
three treatments diﬁ‘ered ﬁ'om the check (Table 16). There were no inter-cultivar
signiﬁcant diﬁ‘erences.

Greenhouse study. There were three cultivars that had statistically signiﬁcant
intra-cultivar diﬁ‘erences. The cv. Superior had signiﬁcant diﬁ‘erences between the check
(36.6) and VD only (19.1) and between the PP/VD treatment (33.5) and VD. For Russet
Burbank, the VD treatment (36.6) differed ﬁom the PPND (27.2). In addition, high PP
(37.2) and PPND (36.1) did signiﬁcantly better than the low PP in the cv. Atlantic (22.5).
As for inter-cultivar comparisons, the only statistical difference was in the VD treatment.
Here, cvs. Russet Burbank (39.9), Snowden (34.9) and Atlantic (33.9) did statistically

better than Superior (19.1).

102

Montcalm research site. The only statistically signiﬁcant difference among
treatments within cultivars was for the cv. Desiree. The VD only treatment (35.9) was
diﬁ‘erent from the check (19.8). Among cultivars, there were some diﬁ‘erences in the
check and PPND treatments, but no diﬁ‘erences among the cultivars in the low PP, high
PP and VD only treatments. Within the check, both Kennebec (35.9) and Hudson (36.9)
responded better than Red Dale (18.9) and Desiree (19.8). Rosa (33.6) was statistically
diﬁ‘erent from Desiree (17.8) in the PP/VD treatment. 9

W. Another method of determining yield loss is to examine how the
treatments diﬁ‘ered ﬁom the check. Because all of the treatments had a low p—value for
replication in yield measurements (Table 1), it was decided to use the diﬁ‘erence between
the treatment and each individual check for the yield loss statistic. Here, three diﬂ‘erent
statistics will be explored. The ﬁrst, is a series of one-sample t-tests that will determine
how different the check is ﬁom each of the treatments. The second will look at intra-
cultivar diﬂ‘erences, and the third will examine inter-cultivar diﬁ‘erences.

Jolly Road site. At the Jolly Road microtile plot, only one treatment with the cv.
Russet Burbank was the yield loss statistically diﬁ‘erent from 0.0 at the p-value <=0.5
level. There were no statistical diﬁ‘erences among treatments within each individual
cultivar. However, when examining inter-cultivar diﬂ‘erences, it was found that in the low
PP treatment, the cv. Russet Burbank (+31.7 g/plant) had a more signiﬁcant loss of yield
than Desiree (~138.5 g/plant). In the VD treatment, Desiree (-202.4 g/plant) was
statistically diﬁ‘erent ﬁ'om every other cultivar except Hudson (-108.2 g/plant). There

were no signiﬁcant diﬁ‘erences in the PPND treatment.

103

Greenhouse study. There were no statistically signiﬁcant diﬁ‘erences in the
greenhouse for any of the statistics.

Montcalm research site. At Montcalm, there were statistical differences between
the treatment and the check for Red Dale low PP and high PP. However, in both cases
the treatment did better than the check (Table 19). The cv. Rosa had statistically
signiﬁcant intra-cultivar diﬁ'erences. The low PP was diﬁ‘erent from the other three
treatments. The only statistical diﬁ‘erence between cultivars in the same treatment was

with the low PP as well. Here again, cv. Rosa was diﬁ'erent ﬁom the rest.

DISCUSSION

Comparing the ten cultivars, it was necessary to assess their relative susceptibility
to Potato Early Die. This was done through relative yields, ranking, yield loss, and
assessing goodness of ﬁt of regression lines. The cultivars were rated as being susceptible
(H), tolerant (M), or resistant (L). Since, even single cultivar potato early die research
results tend to be highly variable, it is not surprising to ﬁnd a high amount of variability
within these ten cultivars. However, it is necessary to examine the trends that develop
rather than the statistically signiﬁcant overall diﬂ‘erences among the cultivars.

The three statistics (relative yield, ranking and yield loss) have a high degree of
variability within their results (Table 22). Each test statistic has associated strengths and
weaknesses. Relative yield uses actual yield data, however, some disparities are a
reﬂection of true differences among the treatments and others are a function of the range.
For example, if a replication has one unusually small yield and another unusually large

yield, it might produce a wide range of values between 0 and 1 that is due largely to

104

random variability within the speciﬁc cultivar rather than true diﬁ‘erences among diﬂ'erent
cultivars. Ranking the yields ﬁ'om small to large will reduce this random error, but it will
also reduce it if the diﬂ'erence is from treatment rather than natural variation. Perhaps,
yield loss is a good estimate in this particular case because there were distinct diﬁ‘erences
in replication (Table 1). However, it is still susceptible to a great deal of inﬂuence ﬁom an
outlier, particularly if the outlier is in the control treatment.

Hence, to get an overall understanding of the relationship among the cultivars
concerning their relative impacts or susceptibility to potato early die a test statistic was
developed that summed the responses from all of the analyses. For each cultivar, the
number of H, M, and L were counted and summed together on the basis of each H
receiving 3 points, M 2 points and a L 1 point. Only the best ﬁt data was used for the
regressions. Hence there were 8 separate assessments for each cultivar (two for relative
yield, two for ranking, two for yield loss, and two for the regression lines).

In summary, Hudson, Russet Burbank, Snowden and Superior were most
susceptible to potato early die. Tolerant varieties included Rosa, Red Dale, and
Kennebec. The most resistant cultivars were Atlantic, Norkota russet and Desiree.

Comparing this analysis to published literature was diﬁcult because most potato
early die studies are done with a single cultivar with various levels of PP and VD both
alone and in combinations. The single most prevalent cultivar for the study of this disease
cycle is cv. Superior (Botseas & Rowe 1994; Francl et al. 1987a, 1987b, 1990; Kotcon &
Loria 1986; Martin et al. 1982; Riedel & Rowe 1985; Rowe et al. 1984, 1987; Wheeler et
a1. 1992, 1994; Wheeler & Riedel 1994). The second most studied cultivar is Russet

Burbank (Davis 1985, Johnson 1988, 1992; Kotcon & Rouse 1984; Kotcon et al. 1984,

105

1985; MacGuidwin 1990, Nicot & Rouse 1987, Rouse 1985). Ohio State University has
done the most published research on cv. Superior, while University of Wisconsin has used
the cv. Russet Burbank. Although the speciﬁc ﬁndings were not as conclusive as the Ohio

State study, Superior was ranking as one of the most susceptible cultivars.

106

Table 1. Three way analysis of variance p-values for variation due to cultivar, treatment and replication for
Wmmmmmmmmddiﬁmtmmmw
quantiﬁcation ofﬁglmmzmmmberyields atthreeexperimental sites.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AN OVA Experiment Cultivar Treatment Replication
Site
Jolly Road 0.001 0.000 0.481
Manchu W
0.967 0.000 0.000
m
Montcalm 0.186 0.000 0.149
Jolly Road 0.001 0.651 0.041
mum. Greenhouse 0. 139 0.080 0.001
rlalllus
Montcalm 0.180 0.396 0.000
JollyRoad 0.025 0.047 0.011
Tuber weight Greenhouse 0.000 0.650 0.000
Montcalm 0.000 0.254 ' 0.000
Jolly Road 0.000 0.034 0.004
Relative
I 1 weight Greenhouse 0.000 0.034 0.000
Montcalm 0.000 0.256 0.000
Jolly Road 0.994 0.21 1 0.008
“Fm“ Greenhouse 0.825 0.052 0.000
by W618!It
Montcalm 0.065 0.359 0.010
Jolly Road 0.015 0.642 ""
Yield loss Greenhouse 0.283 0.437 9'”
Montcalm 0.238 0.554 . "W

 

 

 

 

 

 

107

.8288 8.88 use: a .u Le 85 as: E: ”£98.... 8.88 race 83.8 .u 8 nose. 32 ”e5 8805
98858 as. 52.: 8. 858 gauge» d> nausea 852085 E 888 80.338 games» as a 38885 0.0 “sez

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

g «o: Sandman—.500:

.28 =8 8 88.3.82.
0: a: a: a: o: Md o.o Nd: Wm o.o 2..— od boa w; o=§=<
s: a: 0: :0 a: ma— od «to: Wm N6 N2 m.m eon NA giocm
mo o.o #6 no 0.0 s: :0 a: :0 :0 ho o.o 0mm wd swam
o.o o.o . ON —.0 _.o :0 a: s: 8.... :0 wow o.oom mg 0.2 81:89
o.o o.o Md o.o o.o a: s: :0 0: :0 om v.0 hmm o.o :83:
a: s: s: 0: a: Q: o.o he: _.m _.o vw we we _.o gm Social
0: 0: 0: s: 0!. 5.0 Nd www— o.m o m: 9: non o.o Meant—é «amend
a: :0 0: 0: s: _.mN _.o 0.0: Nd _.o be men 0mm 06— nor—0&5
no _.o me 2 o.o s: s: 0: so... .3... mg ﬁn ow. o.o cocoa—UM
mo o.o «an mo o.o 8.3 :0 o: s: a: N: 26 2.— o.m 0.09 10¢

Dana D> mm 1 mm A MD Dan—m D> me I .3 A MU DEA—m Q> mm v5
ass 8:82 8:882 is 88:85 8.. Box 32 8,25

 

 

 

 

 

 

Emgémﬁggzﬂeﬁnﬂzﬁuguxaggunﬁgggunaccuséwﬂguxagg
90383358829888883582388.hisseoemeogsgaeggeueﬁz mama...

108

£38.... nag 32.: a a go 292 as E: “£283 02:5 3%.: a 5.. ”.32 32 £5 ”.885
2535a 53 UES: :3 333 9:3 as ﬂamed g E J.958 3.33 53.8» B. a gun o.o usz

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

dog you acacia—.50.:
9: o: 2:. :o .3. mod 3 .n ovd Ne; m: 52 cm»; owv 6mm 039:5.
.5... :9 3. 3.. o: 36 wné ové who ow; we». own b8— ?” conicam
h _ as o v YN m._ :5 5: o: o: :8 mwN chm mm@ :2 3cm
a _ M: m m QM N; n: .2. o: :g 8:. wow mm: hm: mug 28D
a o ow v m v.6 o.o 3.. u: :8 y: :5 02. 2.x «EN :2 c9653
:5 o: o: c: o: 3N ha o and me 0 mn; m9. :v as M: counsm 58.qu
u: o: 9:. q: :5 Nvd 2 N mnd on _ «o.o ooo- ooo Nan a; giant—um «own—am
o: c: o: o: o: 3.5 mm 0 R; an _ mod 5w mac m5 m5 Stun—5m
N h _.w o N _.v _.N :9 g: o: 9: :0 £2 wmn 32 an: 009253—
Wm 50 N._ Em nd 9: o: :8 :9 1: «9. $2 mvw mum uﬁﬂ tom
DEAL Q> an I mm .— ¥U as}: D> an I mm 1— M0 Dan—n— D> .3 MD h
5% so: as: 8.355 an Box 2.2 >25

 

 

 

 

 

 

Shh—5:881amino:Ruiniguﬁguxaggguuﬁgggﬂabugaﬁﬁ
BuxgaugésagésexaaasgaéaﬁAaﬁaseuouigggsgsﬁgvgﬁagﬁﬁgz .33.

109

£235 g 833 a .u go 226. is E: 5.38% 99:5. 3.3V 332 a a. 232 32 #3 388a
“895a as 52%. E. 33% 5535» as dag 3335 E 32.80 93.323 ﬂag as gag—yam ”Mo u3oz

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

688. 6: 8935500.:
:8 8:. .2. :3 .8. #2 o.: Wm— 2: ”3 m3 2.— VON woN ou§=<
:- 8: :8 o: u: oNN 5.3 CON 0.: _.mN SN nOn th van EvBOzm
Sm vow >3. 2 : N3 .3 :o 2... :5 .3. 5m va wNn #2 $0M
82 mooN om: vwﬂ NZ“. .: :o q: :5 :9 NVN Sn OMN SN 00:80
No: 82 $2 NV: 03— t: :o o: :- 5: N_N own QMN 3N Genus:
o: c: g: .8. g: NdN _.ON WMN 00— oéN 8n CNN MNN in Ed SStOZ
o: :o o: c: 0: We N.n «N N.v mN 2N NMN 2N mmm xﬁﬁsm gm
2: u: :9 .u. I: «:2 0.: Num— h.m_ _.MN wnN 8N OmN omN how—0&6
0&2 EbN SON woNN SVN o: o: a: :5 9: SN hoN w _ n men 8505—3—
aw: cm: on: we: 32 o: g: u: :8 a: omN VON 5m _Nn 01D voy—
Q>>E D> an I mm A MD D235 D> an I .5 4 MO Dana D> mm Mo .3
ciﬁv a; so; 9593 is» 884895 eﬁgv 2: Box as >25

 

 

 

 

 

 

Emggmgzuﬂﬁggﬁgw
xaﬁﬁnaguuaﬁﬁgéshﬁﬁasguxgéiésasgaﬁegaggaéapﬂgesﬁﬁuﬁaas.‘ .33:

110

 

 

 

 

 

 

 

 

 

 

 

Table 5. Two way analysis of variance (treatment and replication) p-valuea for source of variation due
totreatmentforeachcultivaratthethree experimental sites (JollyRoad, the Greenhouse and
Montcalm Potato Research Farm).
Cultivar Jolly Road Greenhouse Montcalm

Red Dale 0.587 W" 0.124

Kemebec 0.754 "W 0.495

Superior 0.897 0.102 “"

Russet Burbank 0.037 0.549 ""'

Norkota Russett 0.744 0.324 *”

Hudson 0.200 “'" 0.710

Desiree 0.624 ”" 0.292

Rosa 0.844 “* 0.144

Snowden 0.557 0.314 ""”

Atlantic 0.471 0.239 ”“

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4.80: 0.228 o.o 3% 0.89. o.o as 38 o.o n 35.2
”.8me 983% o.o 2: $9.” o.o at. 39. o.o a 53%
ion: ”8&8 o.o 5.: 33m o.o 3m 38 o.o mm «Sm
SEQ 9833” o.o v.32 0.8:~ on... 2: 0.9.: o.o a 8:3
0.35 ”.332: o.o 34: 0.82 o.o Em 38 o.o 2 :83:
323 8:8" o.o 3: ~82 o.o w: E: o.o mm :83 39:02
@805 a. 5 _ E 3 m. E o.o?“ he 2.. SR o.o : 0.5me gm
382 0.88:. o.o I: 98% o.o 2:. 2.8 o.o 8 559m
«.882 $882 o.o «.82 ~89. can as $8 o.o 2 8893.
$38 385 o.o $3 088 o.o Em $9. o.o 8 0.5 8m
:32 S532 S523 so: Essa: SE52 :8: .5522 SEES
95 0:353 gag a $25

 

 

 

 

 

 

§>€sssas ga:»=sa€3~£&é8§.§.€ﬁ gasgaééauﬁm 693:.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2. as o.o 3. o: o.o S q: o.o an «Sm
3 22 o.o 2 in o.o o.o on o.o ; 8:80
2.. an... o.o an ”.2 o.o 3 o.” o.o 8 :83:
3 ed. 2. c.“ ”an 2. 2 E o.o a. 83:60.
2 o.o... o.o S 0.2 o.o 2 q: o.o 9. 25 8m
:82 ea: a: :82 a: a: so: a: 5:
Es: 933% Egg a 525

 

112

.2528 32 2a 55 $5 .5382 E 83.55 is. B. Emoaa ca 35.8 Be an as”. .w as:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o. 2 deN o.o m; YMN o.o NdN 0.8m od Vm 0.28.2
m
WEN .VVNE o.o QN v.9. o.o own 0.3m o.o S covBocm
n.m_ oNoN o.o : QM: o.o NNN 9me o.o G 5833 Sou—.52
v.2 Qwon o.o Q— 92 o.o _.oN o.an o.o em x5925 «085M
Nd—N 0. $5. o.o «an vdn o.o wév 0.8m o.o Nv .5596
E:
502 .5032 53552 :32 8:52.: 8:832 :82 :55: 8:53:
92mm 33% 5:3 a 3% a 55:6

 

 

 

 

 

 

.ggsesemguuaggbugggnagameggtmg #03:.

113

458%.

”22353 ”gauze-:3. . Eda—EN. . . AS §§Hm3§§8a.§§a§§§§ o.o ”3oz
. .Eduail$9739.83 388 «E.

 

mead

mead

coed

Smd

wood

3:.

awed

ooo.o

 

Sod

ooo.o

EON—N

god

ooo.o.

~86 .

N86

856»—

was:

 

~36

. ~36

ooo.o

Sod

and

wood

wand

ooo.o

 

58.0

686

wag:

cmmd

a:

H; «$0.956.

.._. a 35

"advﬂﬂ

£38m

 

Ead

Sad

ooo.o

and

mhvd

oacd

emad

coed

 

ooo.o

ooo.o

avwdnm

«:6

_,_w_m ._ 2.5.”. $3.

a...” m 82.

2:8

38¢

 

cad

62d

5.5

Dn—5

N55

3m...

vaD

95.?

 

mood

ooo.o.

E

 

2N6

:Nd

5.5

LE5

55.

NND5

nn—d-

955?

BE

 

hVN5

a»;

NN05

h:.D

5.5

 

Nmod

ooo.o.

ESCN

mm— 5

55

E55.

 

:md

2.00

5.:

E5

E5

3N5

Eh?

LE

 

R55

 

N56

ooo.o

. SEEN

\ ~55

:55

NB?

A395.

325:.

 

owed

63.0

5.5

5—N5

n.5,:

D—n5

35.5

 

£26

.86.

.5 5mm

Nh—5

:55-

b3:

#5.:

 

3.6

$26

55

Nn—5

 

mood

ooo.o.

23%

33.5

. .53,
.M.m....w..._~3.om

ES

: ...w.~.__,.,=,_. _.

 

 

:nd

:2.

5.5

3N5

. p .35:

 

36.0

08.0.

bong

On—.D

his... .
mgduv

a... w . .3: s.

, . awed .

 

 

v9.0

#96

5.5

n\ 0.5

F

 

wNod

coed

an .me

335

255

L3,...

E5

  

 

 

£55

:b.p

. :N.\nN

 

n.—

 

Pi:

 

380

C

 

92mm

 

 

:5 an» 8:895

!

 

 

 

 

a.

.8380

 

 

Ilnl>
SEE

 

 

.5530

 

 

Agassﬁﬁa paéqausﬁaﬁéseiéaoteaégggg

Eaaﬁaasoé:oE..+§a§-2i£éasE>§+a>+ﬁ+§a§u2ouan§§i§8§338§< «oz:

114

nﬁwoeaq 53 52%— can “353. 35.53 “as g Snag ”.5 ”—258 8.5.5555 55:53 a a gag ”MU

that

.8865
“202

«So u a .33 um 6.93.2.9. + 3:8 u .3.

 

Sad

and

ooo.o

weed

Gad

>36

oewd

V86

 

~56

8&5-

wwvéNN

hN—d

was.

.82

as

$8:

02975

3

0555.4. .

 

n56

mhwd

ooo.o

ovod

~36

Sod

3nd

n—Nd

0227.5

 

Sod

mnmd.

Sndmn

vad

53.2

2.0.2“

nnmdv.

wwwdﬁ

E

.§38m

 

vad

«med

ooo.o

vwvd

Sod

e26

2nd

890

 

«86

8nd-

unaﬁmm

maod

25.2-

Sada

«3.3

Nomdou

0351.

3

8o."—

 

52.9

had

ooo.o

Sod

206

N36

mood

Sod

 

God

35.?

muvdcm

bend

4.52.8“

a:

as:

G 223

02.5-3

E

8.539

 

nmod

nmod

ooo.o

nN—d

mmmd

and

N86

0221.

 

maﬁa

2.56-

3°59”

536

a:

w «:3:

N .5...H.555.mm

2.3:-

3

con—BI

 

and

and

866

and

«26

Ommd

ooo.o

02279

 

«mod

”and

«2..th

>26

M”,3...“me

$2

. 2.3 ‘. .
., 33:

3

32:30.82

 

owed

owed

coed

moud

Onnd

mnnd

oaod

 

8:.

Va _ .a.

wvudou

nm—d

«83

m”.5“.mew..,,

:22

255.9

E

gash—mg

 

hand

and

80.0

02.6

ommd

Sad

Sod

3.97m

 

n86

«and

8903

Need

enhm _

8v.—

3 ~53

E

85.5

 

vmwd

vned

ooo.o

306

God

«had

mend

 

aood

Sud-

nwvd ~ n

owed

«wad.

awaén

«5.3—

255.9

3

 

mad

3:.

ooo.o

and

numd.

and

966

 

awed

SUNS

mmmﬂcu

2.— .o

was

q 1 . <

Hangaa.53...

3.975

3

«in;

 

m

 

Damn

 

ESQ

 

D>

 

 

_. .380

 

 

 

€033... 8:883

 

83853.?—

 

«8m

 

:53

 

 

_ .938.
agéﬁanEgagaﬁésviomzﬁéaésaégﬁésﬁuﬁﬁé:65?+§§u2%3é

3AAoZmes+E>E+ES+§u§u§aB§8§E§§a§§§v§§§25§38§< .223.

115

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

438.5
.6538 ”Etnaﬂggagugaaﬂuaama2858aﬁﬁagggg 06 682
2:. RS 83 and :3 $8 «8... 8o... 3?&
:2 a: .22 3:   H a: :S a; as .3 as:
8:. 3: .o ooo.o :8 .23 , . :3 . _ 53 . 83 8?...
«85 so... «3.: ”So ooo.o :2 :3. 35.8 £8 53%
a; 28 8o... :8 83 23 83 ooo.o 3:1 in:
33 $2. 82: . «8.: 22. ~35 NS... 82: «Be .382
83 83 «8° 89° . 38 $8 8.3 Sod 3.3.“.
£2. :3 3.2 cs... 22. 88 «8.... 3M: «80 383mg
88 33 8o... 23 £2 8:. 83 ooo.o 8:3
”Sc ~86 33.. mod 83 .___._R.~,.o .......(“Bowamwwﬂ, _ _ m8: “.80 8.59m
x 92“: Ease q. 92.: n> .a go
2888 8:895 8.853. :3. 3m :33
.Eouoaauggﬁéagaéiﬁgacaﬁgégg?
£38889.EETaausouEiBéEEETQ:t+§u§u2iba§ inﬁgégﬁsﬁ,‘ .: oz:

 

458E
”839253 52$? 33% as gag E 38:8 §.§§§§§ 06 .902

 

mend mead coed ”mod w—vd «006 35.0 806 313$

 

Sod ash.— amc.o_ n86 VNNA o— _.o . m — ad stum— uooo 0353‘

 

116

we .6 ma: .0 coed m: .o . 5nd 8nd and coed 331m

 

3: a: as a... at? 3:35 .3 3%”

 

82. Sad ooo.o wand chad Sad E.— .c coed 92!:a

 

e8... 2.: 8:: Sod 88.2; 035%., 8o: “80 Ed 3362

 

w _ Nd w _ Nd wmod 8N6 2&6 «2 .c Rho movd 032:—

 

? a: a: a: 2; a: as :2 .3 a3 .3.

V..,..... n...k.>

 

and and coed 936 2.5.6 onmd 3nd coed 0221.

 

w—od chmd v3.2 «.36 8nd. 3%.— nmmd ”8.: “80 in

 

 

 

 

 

 

 

«- 92mm 5:00 a. 92mm D> mm «8560

 

 

€033 8355 8g :5 3m 32.3

 

 

 

 

 

.38.
.uugggpaésﬂﬁaﬁéﬁﬁﬂigéggéuﬁﬁnagegﬁséEa>§§+§§uvwaua§
wean—ZmSigvsiﬁitﬁuauuzﬁﬁa:8§E§gégagi§8§§38§< .223

117

 

4583

”quasi as 52.: 8. ”8:3 5535» Ha> a g E 38:8 3.33 gang 2. a «3335 06 ”3oz

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:. a: 3. Sta :3 23 38 8a.: 8:3

2.5 o: 2.3 shod v2.6 Naad 7 5a."? Sewn» H80 30m

1.... o: 2: Sad 92.6 N36 366 coed 31>-a

o: o: :9 v86 2 in a: .v. 08.2”- 3982 H80 8.489

t. z. z. 23 83 m8... 8% ooo.o 31>...

:. .2 :. RS 82.. ~38 033 $52: and 53m

2: z. 5 Ed 83 a; :3 ooo.o 81>.“

5 :. :. v8... 83.- :3- 33. 22.8" $8 8988M

3. z. :5 E... 83 22. was 82. 31>...

.5. 5 .: Ed 82 32m 803. «222 $8 033.

q. 92% 380 m Es: n> .E 380

gave 835.... 8385i .3... 3m 563
.Agﬁaﬁpéaﬂwguﬁgaﬁésw§§2§§§§§3£§

guaoﬁséxnszasouiﬁéuﬁazmﬁgliﬁusougﬁﬁa:ssoauwaﬁasggﬁai< .2 us:

118

458E

2305“; 53 ”DEE in 0% ESE AS a 3% “mm ”.858 898% ESE as. a 96335 06

”302

 

owmd

$36

3nd

806

8nd

wand

:md

ooo.o

31>.m

 

_ 86

«3.00.

202‘-

gamma

mood

MWa:

5:"

“saga

«000

30M

 

wwmd

wand

sad

ooo.o

v36

auwd

aowd

oood

31:“

 

Sod

mun.”—

8N6

$3.me—

Sod

2:.

snags:

H80

99:39

 

wavd

Sud

wand

ooo.o

god

586

wnwd

ooo.o

nigh

 

68.0

586$

«Gan-

www.35—

maﬁa

Ea

3:.

£5:

«80

8352

 

mbvd

amud

«and

ooo.o

o—wd

«Ned

306

98.0

31>.m

 

N86

.23:

h. .m

S .

232m

Good

2.9.“...

aowdm-

gage?

H30

8388“

 

«3.9

:cd

3N6

coed

Sad

«36

and

68.0

31?.“

 

3:.

253:.

3:.

«8.3

«maﬁa-

Sud—N—

H80

o-aO tom

 

«.—

 

D>

 

mm

 

35350

n.—

 

Q>

 

mm.

 

g0

 

 

9c 81289...

 

8.58 come—Em .85 .23 8.68.53—

 

5m

 

 

525

 

.asaaauﬂueg
25.? 35:53 saggégseoagﬁgﬁ £3688“ 591+Am3+§u§u2033§
Sggusgaéaa 6>+ﬁ+§i€ouaoabnésaamoaaggzgzﬁuﬁsggﬁg<

.3 033—.

119

£2.82. 99...... 33.. an... .... .o 93. a... U...... .233... 99...... 322.. "59.2 .u .o H.2... so. 9......52.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

names... 53 52.... B. 5...... 5.9.9» A... 5 5.995 n... .380 895...... 5.9.9» 9.. .59.. 599.... a... .202

.33. .2. 8955.9:
.: .: :. .: :s E... 8.... 2.... on... 8.... 2.... 9...... 8m... 8...... 2.5....
t. :. :. :. .: ca... 8.... 8...... 8n... 9...... R... E... 8.... 3...... 83.5
E... .2... 9.... .2... an... .: .: .2. t. :. in... a... .3... $2. 33.
9.... N3... 6.... 2.... 3.2. t; :. :. .: :. 5.... $2. 2.... 8.... 8.8m.
3.... 2.... 8.... 8.... 2.... t. t. z. s: :. an... an... an... .9... =89:
.: a: t; .: s: an... an... an... E... 8.... S... :m... :m... an... :88... 53.82
:. :. .: :. t. 2.... 2.... 3...... 8. .. 9.... 9n... 2...... 8.... .8... 95m .33..
:. :. :. .: .: .2... a... 2.... E... a... 8m... 2.... an... 3.... 59.9
wmnd and and mnvd mmvd :3. o: :5 :o 0: 03.0 avmd Nmod cued 00505—3—
936 womd wnvd onmd coed 2: 1: a: o: :5 EVA. w _ v.0 whvd mend o—aﬂ com

9.... E. E .. n... . v... 9.... E. n... = A... .. V... at... n> .. .0 E.
.3.» 29.. .3... 885.8... .9... 2.38.... 25
.83.. 658.. 95.... 61282
. .9 022895 o... .93. 2.2. 8...h 9.2.9....» B... o... .. 5......» 98 a. .22. .3... 89.2... .5958 523.3... .a. -5... .m. 0......

.238... 9.5... 83.. 3.9.... .... 9 H9.... 9... E: .2212... 0...... 30.2.. .899. a... 226. so. in... .88...
9.2.... 53 52.... 9. 3...... 59.9» as 399.. 59995 x... .28.. 8.5.... 5.9.9» a... .55.... 9.999.... Uv... ”2.2

 

 

 

 

120

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.38. .0: 55:94:50.:
2.... 1: :5 t: :8 Son adm «Sm WNN wdm 0.3 Wu. m6. 0.5 . 05:23.
:3 .1... .9: :8 :5 Fan odm Ndm Wow can 5.: 5.2 wdm a.w_ novBoam
Q: OK 0. .N ndm Man :3 .3... i: 3;. 23 ad— nd. QNN We. 3.0%
0.?” ohm mdn wwn wd. 2:. :9 :3. o: .3. .6. o._N m6. v.2 8.995
ohm WON 9mm ném o.o.“ :9 2:. o: .5... .2. v.2 mSn m6. 9: €9.33
.8. u: :3 .3... 2: m.Nm mdm «.mm n.0N Non Na @0— nd. 5. - 309.3. Sow—.52
:5 :9 2.... c: :3 NSN odm 06m ENM 0.; ”2 aw— Qw— wdu SEQ—5m gm
«2.. :o .3 :c :5 Wmm ﬁa— wdm ash 0.9.” SE o.om N: W: Siam
ch OWN nén o.on Qmm on... a: .8. o: :5 Q: Tm. 0.2 mg.— 02.05.3—
FMN 9am o.om mdN ad. 8: :0 q: .3... .9. ww— m.m_ rea— vd. 09D vow—
Q>>E D> mn— : .E .— MD as}: Q> mm 2 .E A MD ERA Q> n..— M0
.9... 3... .3. 8.5.905 .9... 2.82.2 55.5

 

 

 

 

 

.Qwﬁggﬁs.a§a§8m§§..igaa%§aaaéa .2 2....

121

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.... mm on... o.o 8.... S- 35......
2.3 w. 8.... mo 2.... o? .3305
88 .18.- 8.5 c:- ovg 3..- as.
o.o... .2- 88 v.8.- o...o 3..- 8......
88 m..- 0:... 3o.- 88 2.- :88:
SS 3. 8.... 2.. 88 1. :33. 59.52
8.... 2.. $8 3. .3... E. .53.... .83..
9a... m..- 80.0 «.9 2.... 0.. 8.5.5
an... .2 as... v. on... 3.- 8.8.3.
.3... I..- 88 own- 8:. no? 0.5 3.
02.3-.. “Sis 31>-.. .533 .035-.. .533
55.5
g 92.... g 9 A... .E 33
a... o§3§£§a§§.§.§§§a§ .3... 33...... .2 2......

122

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a... o... 8.... m... 88 3.. on... 3...- .8:
2.8 n.2- 9...o So. 28 E.- 83 S.- 02.80
8:. S o8... n... on... E? 88 n8- 8%....
one... Q:- 83 v.2:- oad no..- 88 on..- 8.2.8.
8. .o o... .8... no... .8... o:- muod x... u.5 .3.
31>-.. ESQ» 31>... 21...} 31>-.. .53» 31>-.. 21.9%
g 92.... .8. o> 3.. .E 5.: g .E 33 55.5
.3.» .382 3:382 2.. a .380 o... .9 35.58.. .8. o... a. 398° .8. so... 3.395... E. .3... .35.... .2 2....
o.o.o 3.9 8.... 8.. 88 m..- 2.3 a: 2.5.2
8.... 2... 8...... .3 2.3 8.. EN... .3 832m
8.... m... 8.... e... on... 3... 8.... o... 3.8.. .82...
88 8.. oz... 3..- 88 3o. 8.... 2..- .895 3..
83 8... 8. .o a... o. S a... 83 8.. 8.53
31>-.. E35 31>-.. :13} 31>... Egg 31>-.. «5:5
.3 92.... 3.. o> Ge .5 a... G». .E 33 55.5

 

 

 

 

 

 

.ﬁﬁﬁiﬁaaéaaﬁﬁsssiggﬁuﬁuaégg .2 oz.»

123

£28.... 0.3.... 02...... a .u .o ..25. 5... E: .9652... onus... 30...... a .... .o 226. 3... H1.... .88...
9.0.2.2. 5.... d2... 3.. .33... 53 d> aauaa $5515 u.... ...8...8 81.2.5... gag a... a gang o.o ”soz

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

638. 6: 829800....
9: a: 8: a: 2.... m. m. .m— m. N. w o o m 035.3.
2:. g: 9: 2: :1. n. n. m. n. N. a o m a EvBocm
o. w m o. o. a: o: 2: c: :3 a m ¢ a 33.
m. : N. N. o. :5 o: .8... g: c: w a ¢ a Bar-0D
N. w o. o o. 1.... i: o: 9: u: w w a a :89...
.2. 8: e: 2.... 8: m. n. m. N. N. a o m a 509-5.". 89:02
ii :9 2... 2... 2: m. n. n. m. N. a o a a gig-Sm «ow-ms.
8.... .8... 3.... :0 .5... : : N. n. o. w w a o hem-595
: m. m. I m. .2. u: o: :3. :5 w m w w 00905.3.
N. o. a w a n: 5: o: :0 3.... a w w a 01D to."—
D>\.... 9 ha. I a... A MD as}... E a... E n... A v.0 92.... D> A... MD
81 £888.. 5188.2 .33» 859.595 81 .13. £3 12:20

 

 

 

 

 

 

3.3.598 a... 3.... o... .998. 33.5....» 22.. £8 a. .8312... 3.52 .8 2.3

£285 gage... 335 a .u .0 232 a: E: .222: gage 323. a 5° :32 32 a5 35.2;
“8353 53 ”ESE B. 3...... ﬂag d> a 33335 E :38”. 3.3.3 gang» 2. a gas“ o.o ”3oz

 

 

 

 

124

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.33. 8: 8.35980»:
3. 8. .8 .8 .3 2 N . N . N . .3: w e a a u. , 055.3.
:5 o: :5 u: 2.... N. N. N. N. o: a o o :3 EvBOnw
o v o o :3. :5 o: .3. o: 8: a ¢ o :3. «may.
0 h w w 8: 2:. :9 :5 q: a: w a o 8.3 02.on
a w o o u: I... 2.... :5 3. c: w w a .2. is...
2: t: 2.8 o: :5 N. N. N. .. :3 m a a y: 2083. 39.52
o: :3 2: .3... .: N. N. N. N. :8 ¢ a o 2;. gang. «09.3.
2:. :4 9: o: 2.... m w a o. :5 w w a :5 Siam
.. m. m. .. u: :. g: 2:. o: 5: w h w 9: 89283.
w m N. o :4 :3 :9 :9 .3. a: a w w o: 030 toy.
Damn. D> an. .. ma. .. v.0 DZ”: D> n7. 1 ma. ‘. v.0 gin. D> mm v.0
Ea so: has 32:85 as 33. £2 525

 

 

 

 

 

.aﬁaﬁeugioaéusgagsﬂﬁz .595:

688. 8: 85055500....

A u 0......083 $00. 95. 2....

2 n 8.95 0... a. 0.5 0a...
I u 03.30025 .88 03. 2.... “83882 8.. 15
8:02.805 S...

4 u usages... as. so. a:
2n uﬁnéaczgp

 

 

 

125

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:u .....gsoauénc .33. £28...

.... A 2 .... S. A ... A I .... A I .... A A 0.23.2
... I I ...... I I .... A I ... I A ... I I £38m

A ... A .2 .... A I ... A I .... A I .... I 3.3.

A ... I A ... I .2 ... A A ... A A .... A 09.80
I .... I I ... .2 I .... A .2 ... I I .... I =0..sz
.... .2 A .... I A .... I 2 .... E A ... A A $9.03. 39.qu
.... .I 2 ... A .2 ... .2 I .... I 2 ... I S. ashram. «033.
.... A A .... A I .... I .2 ... A I .... 2 I Siam
.2 ... A A .... I A ... I A .... I S. .... A 02.05qu
I , ... I I ... A A ... A I .... .2 A ... 2 OED vow.
I O .. 2 O .. S. O .. 2 G .. 2 O ..

a. 355%.... «— 35535.63 goA v.0; $5.5”. .30; 0.523. “.5530

 

 

 

 

 

 

 

 

§§§§§§§8uaggﬂu§§§§%§m $.05...

BIBLIOGRAPHY

BIBLIOGRAPHY

Bird, G.W. 1971. Inﬂuence of incubation solution on the rate of recovery of mm:
121-admin]: from cotton roots. Journal of Nematology 3: 378-3 85.

Bird, G.W. 1981. Management of plant parasitic nematodes in potato production IN:
Lashomb, JR and Casagrande, R. (Eds) Advances in Potato Pest Management.
Hutchinson Ross Publishing Company, Stroudsburg, Pennsylvania. 288 pp.

Booth, A. 1963. The role of growth substances in the development of stolons, p. 99-113;
IN: J .D. Ivins and FL. Milthrope (Eds) The Growth of the Potato. Butterworths,
London. 328 pp.

Botseas, DD. and RC. Rowe. 1994. Development of potato early dying in response to
infection by two pathotypes of Wm dahliag and co-infection by
mm: mm. Phytopathology 84(3): 27 5-282.

Brodie, BB. 1984. Chapter 6: Nematode Parasites of Potato IN: Nickle, WP. (Ed)
Plant and Insect Nematodes. Marcel Dekker, Inc., New York. 925 pp.

Burpee, LL. and JR Bloom. 1978. The Inﬂuence of Wench” mm on the
incidence and severity of Momma; Wilt of Potato. Journal of Nematology
10(1): 95-999.

Burton, W.G. 1989. The Potato. Third Edition. John Wiley & Sons, New York. 742
PP-

Chapman, H.W. 1958. Tuberization in the potato plant. Physiologia Plantarum 11: 215-
224.

Chen, J. 1995. Feature, Function, and Nature of Emulenghus mum and W
dahlin interactions associated with Sglanum Wm. Dissertation. Michigan
State University, East Lansing, MI. 196 pp.

Davis, JR 1985. Approaches to control of potato early dying caused by ng'gjﬂinm
dahliaﬁ. American Potato Journal 62: 177-185.

126

127

Dwelle, RB. 1985. Chapter Two: Photosynthesis and photoassimilate partitioning, pp.
35-58, IN: Horton, D., and RL. Sawyer, Eds. Potato Physiology, Academic Press.

Francl, L.J., L.V. Madden, RC. Rowe, and RM. Riedel. 1987. Potato yield loss
prediction and discrimination using preplant population densities of Wm
naming and W: W. Phytopathology 77(4): 579-584.

Francl, L.J., L.V. Madden, RC. Rowe, and RM. Riedel. 1990. Correlation and growing
season environmental variables and the eﬁ‘ect of early dying on potato yields.
Phytopathology 80(4): 425-432.

Francl, L.J., RC. Rowe, RM. Riedel and L.V. Madden. 1988. Eﬁ‘ects of three
soil types on potato early dying disease and associated yield reduction.
Phytopathology 78(2): 159—166.

Gandar, P.W. and CB. Tanner. 1976. Leaf growth, tuber growth, and water potential in
potatoes. Crop Science 16: 531-539.

Haywood, HE. 1938. Solanacease: Sglanum Wm, p. 514-549, IN: The Structure
of Economic Plants. The Macmillan Co., New York. 674 pp.

Hooker, “U. 1990. The potato, pp. 1-6, IN: The Compendium of Potato Diseases, WJ.
Hooker (Ed), APS Press, St. Paul, Minnesota. 125 pp.

Hughes, MS 1991. All eyes on the very important potato. Smithsonian 22(7): 138-149.

Hulsman, 0.0 1982. Inter-relations of root growth dynamics to epidemiology of root-
invading ﬁrngi. Annual Review ofPhytopathology 20:303-27.

Isaac, I. and J .AC. Harrison. 1968. The symptoms and causal agents of early-dying
disease aluminium wilt) of potatoes. Annuals of Applied Biology 61: 231-244.

Iwama, K., K. Nakaseko, A Isoda. 1981. Relations between root system and tuber yield
in hybrid population of the potato plants. Japanese Journal of Crop Science 50(2):
233-238.

Jenkins, W.F. 1964. A rapid centriﬁral-ﬂotation technique for extracting nematodes from
soil. Plant Disease Reporter 48: 692.

Johnson, KB. 1988. Modeling the inﬂuences of plant infection rate and temperature on

potato foliage and yield losses caused by W Mine. Phytopathology
78(9): 1198-1205.

Johnson, KB. 1992. Evaluation of a mechanistic model that describes potato crop losses
caused by multiple pests. Phytopathology 82(3): 363-369.

128

Johnson, K.B., RI. Conlon, S.S. Adams, D.C. Nelson, D.I. Rouse, and PS. Teng. 1988.
Validation of a simple potato growth model in the North Central United States.
American Potato Journal 65: 27-44.

Kotcon, J .B. and R Loria. 1986. Inﬂuence of Wu: mm on plant growth
and water relations in potato. Journal of Nematology 18(3): 385-392.

Kotcon, J B. and DI. Rouse. 1984. Root deterioration in the potato early dying
syndrome: Causes and eﬂ‘ects of root biomass reductions associated with
colonization by Winn dahﬁae. American Potato Journal 61: 557-568.

Kotcon, J.B., D.I. Rouse, and J.E. Mitchell. 1984. Dynamics of root growth in potato
ﬁelds aﬁ‘ected by the early dying syndrome. Phytopathology 74(4): 462-467.

Kotcon, J.B., D.I. Rouse, and J.E. Mitchell. 1985. Interactions of Wham dahliae,
Collemm'chum Mes, thz' gmnm'a mini, and Wench“: penstzans in the
early dying syndrome of russet burbank potatoes. Phytopathology 75(1) 68-74.

Kratzke, M.G. and J P. Palta 1985. Evidence for the existence of functional roots on
potato tubers and stolons: signiﬁcance in water transport to the tuber. American
Potato Journal 62:227-236.

Kumar, D. and RF. Wareing. 1972. Factors controlling stolon development in the potato
plant. New Phytology 71: 639-648.

Lesczynski, DB. and CB. Tanner. 1976. Seasonal variation of root distribution of
irrigated ﬁeld-grown Russet Burbank potato. American Potato Journal 53: 69-7 8.

MacGuidwin, AE. and DJ. Rouse. 1990. Role of mm m: in the potato
early dying disease of Russet Burbank Potato. Phytopathology 80(10): 1077-1082.

Martin, M.J., RM. Riedel, and RC. Rowe. 1982. Wm dahliae and W
W2 Interactions in the early dying complex of potato in Ohio.
Phytopathology 72(6):640-644.

Milthorpe, FL. 1963. Some aspects of plant growth. An introductory survey, pp. 3-16,
IN: J.D. Ivins and FL. Milthorpe (eds) The Growth of the Potato. Butterworths,
London. 328 pp.

Nicot, RC. and D1. Rouse. 1987. Relationship between soil inoculum density of
W dahliae and systemic colonization of potato stems in commercial ﬁelds
over time. Phytopathology 77(9): 1346-1355.

Olthof, T.H.A 1986. Reaction of six Solanum mbmsmn cultivars to Pmtylenchus
mm. Journal of Nematology 18(1):54-58.

129

Plaisted, RH. 1958. Growth of the potato tuber. Plant Physiology 445-453.

Powelson, ML. 1985. Potato eary dying disease in the Paciﬁc Northwest caused by

mm; W pv. carotovora and E. W pv. atroseptica. American
Potato Journal 62: 173-176.

Riedel, RM., and RC. Rowe. 1985. Lesion nematode involvement in potato early dying
disease. American Potato Journal 62: 163-171.

Rouse, DJ. 1985. Some approaches to prediction of potato early dying disease severity.
Amaican Potato Journal 62: 187-193.

Rowe, RC. 1985. Potato early dying - a serious threat to the potato industry. American
Potato Journal 62:157-161.

Rowe, RC., J.R Davis, M.L. Powelson, D.I. Rouse. 1987. Potato early dying: Causal
agents and management. Plant Disease 71(6): 482-489. g

Rowe, RC. RM. Riedel, and M. J. Martin. 1985. Syngergistic interaction between

ymiﬂlliumdahliaeandmtxlenchuspenmnsinpotato earydyins disease
Phytopathology 75(4): 412-418.

Salaman, RN. 1989. The History and Social Inﬂuence of the Potato. Cambridge
University Press. 685 pp.

Smith, HQ 1965. The morphology of Signalman albstatmm, M. dahliae. and M.
trims. New Zeeland Journal of Agricultural Research 8: 450-78.

Stapleton, James. 1995. Linear Statistical Models. Wiley Press, New York. 449 pp.

Townshend, IL. 1973. Survival ometylenghus mm and P. mims in two Ontario
soils. Nematologica 19: 35-42.

Townshend, J .L. and LR Webber. 1971. Movement of mm: mm and the
moisture characteristics of three Ontario soils. Nematologica 17: 47-57.

van Loon, CD. 1981. The eﬁ‘ect of water stress on potato growth, development, and
yield. American Potato Journal 58: 51-69.

Westra, J.V. and K.J. Boyle. 1991. An Economic Analysis of Crops Grown in Rotation
with Potatoes in Aroosth County, Maine. Maine Agric. Exp. Station, University
of Maine, Winslow Hall, Orono, ME 04469-0163. Bulletin No. 834. 39 pp.

Wheeler, TA, L.V. Madden, RM. Riedel, and RC. Rowe. 1994. Distribution and yield-
loss relations of Mullins: dahliae, mm mm. B. ssrihnen. 2.

130

Ma, and Melgjdggyne hapla in commercial potato ﬁelds. Phytopathology
84(8): 843-852.

Wheeler, T.A, L.V. Madden, RC. Rowe, and RM. Riedel. 1992. Modeling of yield loss

inpotato early dying caused byPranrlenchusnmetransandMerticilliumdahliae.
Journal of Nematology 24(1): 99-102.

Wheeler, TA, and RM. Riedel. 1994. Interactions among hummus Mama. 2.
mm and W We in the potato early dying disease complex.
Journal of Nematology 26(2): 228-234.

Wurr, D.C.E. 1977. Some observations of patterns of tuber formation and growth in the
potato. Potato Research 20: 63-75.