.71: ... .
f 51 . .5 . i.
..!1....é_ 1 ‘ . . .. 4n? .
.Q v1.30» “3‘. Eu. 4..
. wwwfiwfim it.»

.«W—HWW xiv.

 

.3
i .5
. Ito‘E-
.. .4 fl. . ‘ , ; . .
Tifififi
is. o 9%. .
.

 

 

 

. 2.9
5%.»...
Titsfi.»

2 .

w;

e.

’4 I. u I
:5 a». ‘

 

 

 

 

. .
.4 A.

 

FSl9

\/

This is to certify that the

thesis entitled

GENOMIC VARIATION AND CONTROL OF
CLAVIBACTER MICHIGANENSIS SUBSP.
MICHIGANENSIS IN MICHIGAN

presented by

Nicole A. Werner

has been accepted towards fulfillment
of the requirements for

M. . .
S degree in Botany and Plant

 

 

Pathology

M ‘ch jES/melllc {34

Major professor

Date [,1 4, 0'

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

 

 

 

LlBRAi-i‘i’
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuepes-pJS

GENOMIC VARIATION AND CONTROL OF CLA VIBACTER MICHIGANENSIS
SUBSP. MICHIGANENSIS IN MICHIGAN

By

Nicole A. Werner

A THESIS

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

MASTER OF SCIENCE

Department of Botany and Plant Pathology

2001

ABSTRACT

GENOMIC VARIATION AND CONTROL OF CLA VIBA C T ER MICHIGANENSIS
SUBSP. MICHIGANENSIS IN MICHIGAN.

By
Nicole A. Werner

Strains of Clavibacter michiganensis subsp. michiganensis were isolated from
infected tomato plants collected from 14 and 9 Michigan commercial fields in 1997 and
1998, respectively, and subjected to rep-PCR fingerprinting. The most commonly
observed C. m. subsp. michiganensis strain from the southwest region was type C (302).
Avirulent type C strains (16%) were detected in two southwest fields. All 85 strains from
a northeast farm were of type D. In 1997, strains isolated from two fields of one north-
central grower were primarily type C strains (58) with few type D (2). In 1998, two
north-central fields of the same grower contained 35 type D and 9 type C strains. In 1997
and 1998, the tomato fields sampled in the southeast region of Michigan contained 66 A,
63 B and 78 C type strains.

In 1996, copper hydroxide alone and with mancozeb or streptomycin reduced
pathogen populations relative to acibenzolar-S-methyl (ASM), ASM/copper hydroxide
and avirulent C. m. subsp. michiganensis strains. In 1997 only, copper hydroxide/
mancozeb limited spread compared to copper hydroxide/streptomycin. In the field,
inoculated control plants produced yields that were 61% (1996), 93% (1997) and 98%
(1998) of those produced by the uninoculated controls. Fruit spotting occurred regardless

of treatment and was highly variable.

DEDICATION

In loving memory of my grandfather, who instilled in me an interest and respect
for the environment that has remained with me to adulthood. His love for the outdoors

and the living things around him has inspired me to choose a career in biology.

iii

ACKNOWLEDGMENTS

I would like to thank my co-advisor, Dr. Mary Hausbeck, for giving me the
opportunity to study under her guidance and for the work-related as well as moral support
she has afforded me over the years. I would also like to thank my co-advisor, Dr. Dennis
Fulbright, for his guidance and questions he challenged me with over the course of my
studies. In addition, Dr. Andy J arosz has been invaluable with advice on statistical issues

as well as with insightful comments and suggestions.

I couldn’t have achieved my objectives without the assistance of the technical
staff and graduate students of Dr. Hausbeck’s and Dr. Fulbright’s labs. The countless
hours of spray applications, plantings, harvests and data collection could not have been
done without them. In addition, Dr. Robert Podolsky’s statistical consulting was critical

for the completion of this project.

I thank my parents for investing in my undergraduate studies and for being there
for me under any circumstance. And last, but not least, I would like to thank my good
friend, Craig Skoczylas, for supporting me in my studies in addition to uprooting from his

life-long home to accompany me during my graduate career.

iv

TABLE OF CONTENTS

LIST OF TABLES. ..................................................... vi
LIST OF FIGURES ..................................................... viii
LITERATURE REVIEW. ................................................. 1
Introduction 1
The Bacterial Canker Pathogen 2
Molecular Techniques 8
Control Methods 10
Literature Cited 12

I. GENOMIC VARIATION OF CLA VIBAC T ER MICHIGANENSIS SUBSP.

MICHIGANENSIS IN MICHIGAN. .................................. 19
Abstract 20
Introduction 21
Materials and Methods 22
Results 36
Discussion 43
Literature Cited 49

II. EFFECT OF BACTERICIDES, RESISTANT CULTIVARS, AN SAR-
INDUCING COMPOUND AND AVIRULENT STRAINS ON POPULATION
SIZE AND SPREAD OF CLA VIBAC T ER MICHIGANENSIS SUBSP.
MICHIGANENSIS ON SEEDLING TOMATOES IN THE GREENHOUSE. . . 52

Abstract 53
Introduction 54
Materials and Methods 55
Results 67
Discussion 77

Literature Cited 83

Table

II.

LIST OF TABLES

F:
m
(D

GENOMIC VARIATION OF CLA VIBA C TER MICHIGANENSIS SUBSP.
MICHIGANENSIS IN MICHIGAN

Strains of Clavibacter michiganensis subsp. michiganensis isolated
from fruit and foliage collected from Michigan commercial fresh market
and processing tomato fields in 1997 and 1998. ......................... 23

BOX-PCR fingerprint types of Clavibacter michiganensis subsp.

michiganensis and isolated from fruit and foliage collected from

Michigan commercial fresh market and processing tomato fields in 1997

and 1998 according to region. ....................................... 38

BOX-PCR fingerprint types of Clavibacter michiganensis subsp.
michiganensis strains isolated from foliage collected in consecutive years
(1997 and 1998) from Michigan commercial fresh market tomato fields. ..... 4O

BOX-PCR fingerprint types of Clavibacter michiganensis subsp.
michiganensis strains isolated from foliage collected at three times from
a Michigan commercial ‘Mountain Spring’ tomato field in 1997. .......... 41

EFFECT OF BACTERICIDES, RESISTANT CULTIVARS, AN SAR-
INDUCING COMPOUND AND AVIRULENT STRAINS ON POPULATION
SIZE AND SPREAD OF CLA VIBACTER MICHIGANENSIS SUBSP.
MICHIGANENSIS ON SEEDLING TOMATOES IN THE GREENHOUSE

Summary of contrast results comparing treatments applied to tomato

seedlings in the greenhouse when inoculated with Clavibacter

michiganensis subsp. michiganensis in relation to pathogen population

size in 1996. ..................................................... 68

Summary of contrast results comparing treatments applied to tomato

seedlings in the greenhouse when inoculated with Clavibacter

michiganensis subsp. michiganensis in relation to pathogen

size in 1998. ..................................................... 70

vi

Table Page

3. Fruit yield and summary of contrast results comparing
treatments on tomato plants; seedlings inoculated or not inoculated
with C Iavibacter michiganensis subsp. michiganensis and subjected
to spray treatments in the greenhouse in 1996 and 1997. .................. 74

4. Incidence of fruit spotting on tomato plants in the field; seedlings
inoculated or not inoculated with Clavibacter michiganensis subsp.
michiganensis and subjected to spray treatments in the greenhouse in

1996 and 1997. ................................................... 78

vii

Figure

11.

LIST OF FIGURES

GENOMIC VARIATION OF CLA VIBACTER MICHIGANENSIS SUBSP.
MICHIGANENSIS IN MICHIGAN

Four regions of Michigan from which Clavibacter michiganensis subp.
michiganensis strains were isolated from tomato plants and fruit collected from
processing and fresh market commercial fields in 1997 and 1998. ........... 33

EFFECT OF BACTERICIDES, RESISTANT CULTIVARS, AN SAR-
INDUCING COMPOUND AND AVIRULENT STRAINS ON POPULATION
SIZE AND SPREAD OF CLA VIBAC T ER MICHIGANENSIS SUBSP.
MICHIGANENSIS ON SEEDLING TOMATOES IN THE GREENHOUSE

Diagrams of treatment blocks with shaded areas serving as a buffer

between blocks and each rectangle representing a 52.5 x 26.5 cm tray

containing 288 tomato seedlings. In 1996-1997 (A), seedlings were

sampled from each block following the diagonal (regions a to d) from the
Clavibacter michiganensis subsp. michiganensis inoculum focus C“).

In 1998 (B), seedlings were sampled from each block following the

diagonal (regions a to e) between the C. michiganensis subsp.

michiganensis inoculum foci (*) ...................................... 58

C lavibacter michiganensis subsp. michiganensis populations

(CFU/g) on treated or untreated seedlings in the greenhouse

prior to transplanting in the field in 1996 (D) and 1997 (I) when

sampled from sites adjacent to C. michiganensis subsp.

michiganensis inoculum focus (a), the center of the treatment

block (b, c), and the point farthest from the inoculum focus (d).

Error bars represent standard error . . . . 71

viii

Figure Page

3. Clavibacter michiganensis subsp. michiganensis populations(CFU/g) on treated
or untreated tomato seedlings in the greenhouse prior to transplanting in the field
in 1998 when sampled from sites adjacent to C. michiganensis subsp.
michiganensis inoculum foci (a, e), the center of the block (c), and the points in
between (b, d). Error bars represent standard error. ...................... 72

ix

LITERATURE REVIEW

INTRODUCTION

In 1999, the Michigan fresh market tomato (Lycopersicon esculentum Mill.)
industry comprised 2,700 of the 111,750 harvested vegetable acreage in Michigan and
generated nearly 21 million dollars at the point of sale (4). In Michigan, tomato seedlings
are grown in local greenhouses and transplanted into fields for fruit production. Large-
scale producers may grow a single or a few cultivars in their fields, while smaller road-
side market growers may plant several cultivars to satisfy the needs of varied clientele.
Common cultural practices include methyl-bromide fumigation of raised beds covered
with black plastic. The plants are supported by tying stems to wooden stakes. A
customary method for disease management is rotation; the practice of refraining from
growing the same crop in a particular field over consecutive years. Some growers in
Michigan may move tomato production miles away from a previous year’s field while
other growers, having limited options, may rotate to a field within yards of a previous
year’s planting. The number of growing seasons a field is kept out of tomato production
also varies depending on the grower’s situation.

Bacterial canker of tomato, caused by Clavibacter michiganensis subsp.
michiganensis (Smith) Davis et al., is an important disease of tomato in Michigan. A
significant outbreak occurred in 1969 in the north-central and northeast United States
(16). A more recent epidemic occurred in 1984 in the mid-westem United States and

Canada (20). Over the years, the disease has caused significant financial loss for both the

processing and fresh market tomato industries. In Ontario, Canada yield losses averaged
5-10% annually from 1965-1971 (32).

Transplants grown in greenhouses are subjected to humid conditions and overhead
watering conducive to bacterial growth and spread (23). An important source of inoculum
is infested seed (7, 17, 21, 41, 47). Seedlings may become systemically infected yet
remain asymptomatic (l 8) and as a consequence serve as primary inoculum for field
outbreaks. In addition, bacteria are spread throughout the field by various means of
dispersal and mechanical damage resulting in secondary infections.

THE BACTERIAL CAN KER PAT HOGEN

Clavibacter michiganensis subsp. michiganensis was first reported in 1909 by
Erwin F. Smith following detection on greenhouse tomatoes in Grand Rapids, Michigan
(41). Today this organism causes disease in all regions of the world where tomatoes are
grown, and is important in the seed production regions of Asia, Mexico and South
America (41).

Plants associated with the pathogen. The pathogen affects not only tomato, but
other members of the Solanaceae as well. Sweet pepper (Capsicum annuum) (51) and
bell pepper (C. frutescens) (26) are also hosts of C. michiganensis subsp. michiganensis,
although tomato remains the only economically important host (21). The spiny Puerto
Rican weed (Solanum mammosum) (40), common perennial nightshade (S. douglasii) (2),
black nightshade (S. nigrum) (1) and hairy nightshade (S. triflorum) (1) are all hosts of C.
michiganensis subsp. michiganensis, having been found naturally infected near tomato

fields. Further study indicated that an isolate from tomato was pathogenic to tomato, yet

not to S. douglasii (2). Additional Solanaceae species and genera have shown some
response only through seed, root, stem, or cotyledon inoculations (1, 22, 40, 46).

Symptoms on tomato. Symptoms caused by the pathogen can be classified into
two groups: localized, resulting from surface infection; and systemic, resulting from
vascular infection (41). Marginal leaf necrosis, often referred to as the “firing” stage, is a
common symptom in the field. This necrosis tends to be limited to the leaf margins and
may border chlorotic tissue. Local fruit lesions, commonly called bird’s eye spots,
develop under conditions of high relative humidity (41 ). The lesions first appear as
white, superficial spots that eventually develop a necrotic center and are generally 3-6
mm in diameter. Bird’s eye spots are a distinctive symptom of the disease, therefore, are
useful in diagnostics. A recent study found that when inoculated, flowers were most
susceptible to infection two-days post anthesis (34). In addition, small green fruit
developed bird’s eye lesions when inoculated with C. michiganensis subsp.
michiganensis.

One of the first systemic symptoms observed in field-grown plants is wilting,
which generally begins on the basal leaves and continues up the plant. Unique to
bacterial canker, leaves may wilt unilaterally until eventually the entire leaf withers,
although the petiole holds its turgidity for some time. When severely infected, the plant
may die. More often, however, the plant survives, yet yield is notably reduced. Yellow
to brown discoloration may occur along stems and petioles, which may split resulting in
cankers. Less commonly, fruits may be systemically infected through the vascular system

and become stunted, mottled, and malformed (27). Symptom development is likely

affected by a wide range of factors including pathogen variability, environmental
conditions, age of plants when infected, and infection site (41).

In the early stages of systemic infection, bacteria invade the xylem vessels of the
plant. At this time, only the invaded cells are affected, suggesting cellulolytic and
hemicellulolytic enzyme activity (52). Prior to vessel wall degradation, swelling, and
shredding of the walls and middle lamella may occur, suggested to be caused by increased
acidity or cation complexing organic acids (5).

During later stages of infection, Wallis (52) found that the contiguous walls and
middle lamella adjacent to infected vessels are susceptible to degradation in the presence
or absence of bacteria suggesting that the production and translocation of extracellular
bacterial enzymes through conductive tissues. Furthermore, primary vessel walls, middle
lamella, parenchymatous ground tissue, and phloem cells may be invaded and collapse.

Three explanations were proposed for the wilting symptoms caused by bacterial
canker; vessel plugging, toxic action theory (41), and enzyme activity (52). Possible
substances that may block vessels include extracellular polysaccharides (EPS), bacterial
cells, or degraded tissue (13). However, Wallis (52) found little EPS, low concentrations
of bacterial cells, and a minimal amount of degraded tissue in infected stems. A related
pathogen, C. michiganensis subsp. sepidonicus, produces a toxin which causes cell walls
to degrade in a layering pattern (43). Wallis (52) reported that in the presence of C.
michiganensis subsp. michiganensis the vessel walls do not exhibit this layering pattern

and are completely dissolved, suggesting that an enzyme, rather than a toxin, is at work.

Epidemiology of bacterial populations on greenhouse grown transplants.
Clavibacter michiganensis subsp. michiganensis survives on the surface and in the layers
of the seed coat (7, 41) and can be transmitted to emerging seedlings (7, 10, 47). Chang
et a1. (10) investigated the transmission of C. michiganensis subsp. michiganensis from
seed to transplant by growing seed from systemically infected plants and observed a seed
transmission rate of 0 to 0.9%. Bryan (7) found 1 to 5% seed transmission from naturally
infected seeds and 21 to 40% with artificially infested seed. As seedlings emerge,
epiphytic bacteria thrive in the humid greenhouse (47) and may spread to adjacent,
uninfected plants via overhead watering (Hausbeck, unpublished data). In a study
performed by Hausbeck, et a1. (23), populations on greenhouse plants reached levels
greater than 108 CF U/g fresh weight. The study determined that pathogen populations on
greenhouse grown transplants must reach a threshold level of 1.0 x 107 CFU/g of tissue if
severe symptoms and yield losses are to occur in the field.

Spread of pathogen populations among transplants can occur during handling of
both infected and healthy seedlings at the time of transplanting (32). In addition,
seedlings are often pruned prior to transplanting, resulting in the dissemination of the
organism from infected to healthy plants (9). Populations were first detected on pruned,
asymptomatic seedlings 9 to 13 days after clipping. Concentrations of the pathogen
ranged from non-detectable levels to 107 CFU/ g fresh weight throughout the study.

Epidemiology of bacterial populations on tomato plants in the field. Bryan
(7) reported that a 1% seed transmission rate resulted in 54.5% infection of field plants.

Field sources of inoculum include infected transplants, surface pathogen populations on

tomato and alternative weed hosts, and overwintered tomato debris (10). Epiphytic
populations on field plants may spread by water splashing, pesticide spraying, pruning,
tying, sand blasting, and harvesting (20). The highest incidence of disease in the field
occurred when seedlings were clipped prior to transplanting (10). Leaf surface
populations may play a larger part in causing systemic infection of fresh market tomatoes
which are more susceptible to wounding during staking and removal of axillary leaves
(20, 25).

Gleason et al. (19) reported that epiphytic populations on fruit and foliage in an
Iowa research field increased and stabilized at 10" tolO8 and 105t010" CFU/g,
respectively. In a study by Chang, et a]. (10), bacteria remained on uninoculated tomato
plants for 7 weeks after infected neighboring plants were removed. However, population
sizes were significantly smaller than those treatments where infected plants were not
removed, indicating the potential of secondary spread. Secondary infection was not
observed until populations reached 106 and 107 CFU/g fresh weight and appeared 56 to 70
days after transplanting, both depending on susceptibility of the cultivar (10). Ricker and
Reidel (37) determined that secondary spread did not reduce yield of processing tomatoes.
In another study, systemically infected plants exhibited a 5-7% decrease in yield for each
10% of systemically infected plant incidence (l 1).

Alternative host plants may also harbor surface populations (see previous section
“Host Range”). Chang, et a1. (10) found that surface populations on tomato plants (10"-
109 cfu/g) were higher than on alternative host plants (0-108 cfu/g), and were subsequently

higher than on non-host plants (0-103 cfu/ g). Leaf surface populations on alternative host

plants did not increase throughout the season as they did on tomato.

The pathogen overwinters in tomato debris as found in Georgia (8), Utah (6), Italy
(12), California (21), North Carolina (42) and the Midwest United States (15, 19).
Experiments with greenhouse soils (54) and outdoor soil tubes in Washington, D. C. and
New York (7) determined that survival of the pathogen occurred from a few to l 1
months, respectively. Basu (3) found that without debris the organism survives only 3-4
weeks in the soil at 5 to 35°C, yet survives up to 36 weeks in soil at -20°C. Both Chang
et a1. (10) and Gleason et a1. (20) reported a higher rate of bacterial population decline
when C. michiganensis subsp. michiganensis infested tomato debris was buried as
opposed to being left on the soil surface. In an experiment performed by Gleason (20),
populations were detected on foliage of plants afier one week of transplanting into a
debris infested field. After 5-7 weeks, field populations had reached 106-108 CPU/leaflet.
The pathogen has the potential to overwinter for 2 years in the north-central US (19).
There is also evidence that the pathogen overwinters on the weed host S. douglasii (2).

Detection and growth in culture. D2 media selects for C. michiganensis subsp.
michiganensis by altering the surface components of bacterial membranes to select for a
specific genus (24). D2 media was efficient in detecting population levels up to 107 cfu/g
infected tissue and successfully selected for C. michiganensis subsp. michiganensis in a
mixture of other important phytobacteria. F atmi and Schaad (17) developed semi
selective media for C. michiganensis subsp. michiganensis (SCM) that was successful in
detecting one infested seed in 10,000. More importantly, C. michiganensis subsp.

michiganensis colonies were morphologically distinct from most saprophytic

contaminants. Researchers modified SCM to enhance the distinction of C. michiganensis
subsp. michiganensis colonies and reduce the size of saprophytic colony size (5 3).
MOLECULAR TECHNIQUES

Bacterial genomes typically include a variety of low copy number repeated
sequences including interspersed repetitive DNA sequences (IRSs) that range from 15 to
several hundred base pairs (49). They are widely dispersed throughout prokaryotic
genomes, non-coding, and intercistronic. Examples of known conserved DNA elements
include repetitive extragenic palindrome (REP), enterobacterial repetitive intergenic
consensus (ERIC), and interspersed, repetitive BOX sequences (BOX) (49). The
presence, widespread distribution and highly conserved nature of these sequences suggest
that these repetitive sequences are important to the structure and evolution of the bacterial
genome, yet their function(s) is not yet known (31). These elements are thought to be
present in both orientations within bacterial genomes (38) and contain highly conserved
regions that have been utilized as primer binding sites in PCR to amplify DNA sequences
between the elements (49). BOX elements, in particular, are modular in nature, consist of
three differentially conserved subunits, and were the first IRSs found in a Gram +
bacterium (Streptococcus pneumoniae) (49).

Repetitive-sequence-based (rep)-PCR is a method used to identify and
differentiate microbes. The technique utilizes outwardly facing primers complementary
to IRSs within the target genome resulting in the amplification of differently sized DNA
fragments between the elements. When the PCR products are fractionated utilizing gel

electrophoresis, PCR-DNA fingerprints are generated (28, 29, 30, 38, 49, 50).

Rep-PCR differs from other PCR procedures in that it allows the simultaneous
amplification of varied sizes of DNA fragments. Smaller fragments amplify more
efficiently than larger ones, therefore, conditions must be adjusted to allow the
amplification of all fragments (49). In order to accomplish this, excess dNTP’s, primer,
and initial template DNA are added (49). In addition, a longer elongation time during
each cycle is needed. It was later found that reproducibility of REP-fingerprints was
possible using whole-cell PCR (3 8). Because of possible contamination and amplicon
carryover between cycles, a negative control must be run each time (49).

PCR amplification of Gram + bacterial DNA can be challenging because of the
presence of a single lipid bilayer with a thick peptidoglycan layer and an additional
carbohydrate polymer, lipoteichoic acid (35). In order to lyse the Gram positive cells,
detergent and enzymes are included in the PCR reaction mixture to dissolve the lipid
membrane and to cleave the glycosidic linkages, respectively (49). If the cells are lysed
in the reaction tube, the presence of cell wall constituents, cell proteases, and cell
nucleases may interfere with the reaction (49). Versalovic, et al. (49) suggested that by
reducing the preparation time, the loss of Taq DNA polymerase and the original target
DNA may be avoided. In addition, immediately freezing the amplified products may
reduce any loss of amplified product.

Subdivisions within C. mickiganensis subsp. michiganensis distinguished
using BOX-PCR. Louws, et al. (29) reported that using rep-PCR with the BOX primer
(BOX-PCR) C. michiganensis subsp. michiganensis is distinguished into four goups (A,

B, C, D). Utilization of this finding expedites diagnosis and may have useful

implications to epidemiological studies. Epidemiological studies have traditionally
utilized antibiotic resistance as a method to mark strains. Rep-PCR may be employed to
mark strains in epidemiological studies as well as in plant-microbe interactions (28).
Another use of rep-PCR may be to determine the origin of an infection or epidemic (33).

Louws, et al. reported that 30% of the A type strains included in their study were
non-pathogenic (29). Additionally, the type A strains were found to be associated more
frequently with processing-type cultivars than any other PCR type. The majority of the
strains representing Michigan field locations in this study were of the C and B fingerprint
types. The type D was the least represented type and was detected in only two Michigan
fields in the study.

CONTROL METHODS

Bactericidal sprays. Traditionally, growers apply bactericides to field plants
after disease symptoms have appeared. Typically, the products used to control C.
michiganensis subsp. michiganensis are c0pper based products alone or mixed with
mancozeb. A North Carolina study conducted on a fresh market cultivar indicated
significant control of foliar marginal necrosis and a decrease in yield loss with weekly
field applications of copper (39). These treatments did not, however, impact fruit
spotting.

A recent study focused on greenhouse applications of copper alone or with
mancozeb and found the treatments reduced pathogen population size and spread among
greenhouse seedlings and reduced subsequent yield loss in the field (23). They also

determined that bacterial populations on the greenhouse transplants were suppressed

10

below the treshold level (107 CFU/g fresh weight) needed for severe symptoms to occur
in the field.

Resistant cultivars. Resistance to bacterial canker has been reported in two wild
species of tomato, L. hirsutum (22) and L. pinpinefollium (44, 45). Accessions of L.
esculentum with similar phenotypes to the wild species have also been noted (44, 45).
Processing cultivars developed by H. J. Heinz Co., Pittsburgh, PA have held promise in
resistance screenings (14). Tomato cultivars moderately resistant to bacterial canker have
been developed that exhibit reduced foliar blight and yield losses compared to a
susceptible cultivar (36). No tomato species have been found to be immune to the
pathogen. Somaclonal variation was explored by van den Bulk, et al. (48) with limited

resistance potential.

11

10.

LITERATURE CITED
Ark, P. A., and Thompson, J. P. 1960. Additional hosts for tomato canker
organism, Corynebacterium michiganense. Plant Dis. Rep. 44:98-99.
Baines, R. C. 1947. Perennial nightshade, a host for Corynebacterium
michiganense. Report of the twenty-eighth annual meeting of the pacific division
of the American Phytopathological Society. 37:359.
Basu, P. K. 1970. Temperature, an important factor determining survival of
Corynebacterium michiganense in soil. Phytopathology 60:825-827.
Bay, D. M., and Wyant, D. (eds). 1999. Michigan Agricultural Statistics. MDA,
Michigan pp. 42-45.
Beckman, C. H. 1964. Host responses to vascular infection. Ann. Rev. of
Phytopathol. 2:231—252.
Blood, H. L., and Warm, F. B. 1932. Botany and Plant Pathology. Pages 49-54
In: Biennial Rept. Utah Agr. Ext. Sta. 1931/1932. Bull. 235.
Bryan, M. K. 1930. Studies on bacterial canker of tomato. J. Agr. 41:825-851.
Bryan, M. K., and Boyd, 0. C. 1930. Control of bacterial canker of tomatoes.
(Abstr.). Phytopathology 20:127.
Chang, R. J., Ries, S. M., and Pataky, J. K. 1991. Dissemination of Clavibacter
michiganensis subsp. michiganensis by practices used to produce tomato
transplants. Phytopathology 81 : 1276-1281 .
Chang, R. J., Ries, S. M., and Pataky, J. K. 1992. Local sources of CIavibacter

michiganesis subsp. michiganensis in the development of bacterial canker on

12

11.

12.

13.

14.

15.

16.

17.

18.

tomatoes. Phytopathology 82:553-560.

Chang, R. J ., Ries, S. M., and Pataky, J. K. 1992. Reductions in yield of
processing tomatoes and incidence of bacterial canker. Plant Dis. 76:805-809.
Ciccarone, A., and Carill, A. 1948. Field observations on Corynebacterium
michignense (Smith) Jensen and consideration on a possible instance of its
survival in the soil. Boll. Staz. Pat. Veg. Roma Ser. 3. 6:177-179.

Dimond, A. E., and Waggoner, P. E. 1953. On the nature and role of vivitoxins
in plant diseases. Phytopathology 43:229-235.

Emmatty, D. A., and John, C. A. 1973. Evaluation of resistance to bacterial
canker of H2990, a new tomato variety. Plant Dis. Rep. 57:584-586.

Farley, J. D. 1971. Recovery of Corynebacterium michiganense from
overwintered tomato stems by the excised-petiole inoculation method. Plant Dis.
Rep. 55:654-656.

Farley, J. D., and Miller, T. D. 1972. Spread and control of bacterial canker in
tomato transplants during clipping. Ohio Agr. Res. Develop. Center Res. Sum.
65:23-25.

F atmi, M., and Schaad, N. W. 1988. Semiselective agar medium for isolation of
C lavibacter michiganense subsp. michiganense from tomato seed.
Phytopathology 78: 121 -1 26.

Gitaitis, R. D., Beaver, R. W., and Voloudakis, A. E. 1991. Detection of
Clavibacter michiganensis subsp. michiganensis in symptomless tomato

transplants. Plant Dis. 75:834-838.

13

19.

20.

21.

22.

23.

24.

25.

26.

Gleason, M. L., Braun, E. J., Carlton, W. M., and Peterson, R. H. 1991. Survival
and dissemination of Clavibacter michiganensis subsp. michiganensis in
tomatoes. Phytopathology 81 : 15 19-1523.

Gleason, M. L., Gitaitis, R. D., and Ricker, M. D. 1993. Recent progress in
understanding and controlling bacterial canker of tomato in eastern North
America. Plant Dis. 77: 1069-1076.

Grogan, R. G., and Kendrick, J. B. 1953. Seed transmission, mode of
overwintering and spread of bacterial canker of tomato caused by
Corynebacterium michiganense. (Abstr.). Phytopathology 43:473.

Hassan, A. A., D. L. Strider, and Konsler, T. R. 1968. Application of
cotyledonary symptoms in screening for resistance to tomato bacterial canker and
in host range studies. Phytopathology 58:233-239.

Hausbeck, M. K., Bell, J., Medina-Mora, C., Podolsky, R., and Fulbright, D. W.
2000. Effect of bactericides on population sizes and spread of Clavibacter
michiganensis subsp. michiganenesis on tomatoes in the greenhouse and on
disease development and crop yields in the field. Phytopathology 90:38-44.
Kado, C. I., and Heskett, M. G. 1970. Selective media for isolation of
Agrobacterium, Corynebacterium, Erwim'a, Pseudomonas, and Xanthomonas.
Phytopathology 60:969-976.

Kontaxis, D. G. 1962. Leaf trichomes as avenues for infection by
Corynebacterium michiganense. Phytopathology 52: 1306-1307.

Lai, M. 1976. Bacterial canker of bell pepper caused by Corynebacterium

14

27.

28.

29.

30.

31.

32.

33.

34.

michiganense. Plant Dis. Rep. 60:339-342.

Layne, R. E. C. 1967. Foliar trichomes and their importance as infection sites for
Corynebacterium michiganense on tomato. Phytopathology 57:981-985.

Louws, F. J ., Bell, J ., Medina Mora, C. M., Smart, C. D., Opgenorth, D.,
Ishimaru, C. A., Hausbeck, M. K., deBruijn, F. J ., and Fulbright, D. W. 1998.
rep-PCR-mediated genomic fingerprinting: a rapid and effective method to
identify Clavibacter michiganensis. Phytopathology 88:862-868.

Louws, F. J., Fulbright, D. W., Stephens, C. T., and Bruijn, F. J. (1. 1995.
Differentiation of genomic structure by rep-PCR fingerprinting to rapidly classify
Xanthomonas campestris pv. vesicatoria. Phytopathology 85:528-536.

Louws, F. J ., Fulbright, D. W., Stephens, C. T., and De Bruijn, F. J. 1994.
Specific genomic fingerprints of phytopathogenic Xanthomonas and
Pseudomonas pathovars and strains generated with repetitive sequences and PCR.
Appl. Environ. Microbiol. 60:2286-2295.

Lupski, J. R., and Weinstock, G. M. 1992. Short, interspersed repetitive DNA
sequences in prokaryotic genomes. J. Bacteriol. 174:4525-4529.

McKeen, C. D. 1973. Occurrence, epidemiology, and control of bacterial canker
of tomato in southwestern Ontario. Can. Plant Dis. Surv. 53: 127-130.

McManus, P. S., and Jones, A. L. 1995. Genetic fingerprinting of Erwinia
amylovora strains isolated from tree-fi'uit crops and Rubus spp. Phytopathology
85:1547-1553.

Medina-Mora, C. M., Hausbeck, M. K., and Fulbright, D. W. 2000. Bird's eye

15

35.

36.

37.

38.

39.

40.

41.

42.

lesions of tomato fruit produced by aerosol and direct application of Clavibacter
michiganensis subsp. michiganensis. Plant Dis. (In press).

Prescott, L. M., J. P. Harley, and Klein, D. A. 1993. Microbiology, 2nd ed. in K.
Kane, Ed. Wm. C. Brown Publishers, Dubuque, Iowa.

Ricker, M. D., Emmatty, D. A., Schott, M. D., and Ozrninkowski, R. H. 1995.
Heinz tomato varieties resistant to bacterial canker. Pages 37-39 in: 10th Annual
Tomato Disease Workshop, Windsor, Ontario Canada.

Ricker, M. D., and Riedel, R. M. 1993. Effect of secondary spread of
Clavibacter michiganensis subsp. michiganensis on yield of northern processing
tomatoes. Plant Dis. 77:364-366.

Schneider, M., and De Bruijn, F. J. 1996. Rep-PCR mediated genomic
fingerprinting of rhizobia and computer-assisted phylogenetic pattern analysis.
World J. Microbiol. Biotechnol. 12:163-174.

Shoemaker, P. B. 1992. Evaluation of different materials for tomato bacterial
canker, 1991. Fung. Nematicide Tests 47:155.

Smith, E. F. 1914. The Grand Rapids tomato disease. Carnegie Institution,
Washington.

Strider, D. L. 1969. Bacterial canker of tomato caused by Corynebacterium
michiganense; a literature review and bibliography. North Carolina Experiment
Station, Technical Bulletin 193:110 pages.

Strider, D. L., and Konsler, T. R. 1965. Cotyledonary symptoms of bacterial

canker of tomatoes. Plant Dis. Rep. 49:634-635.

16

43.

44.

45.

46.

47.

48.

49.

50.

51.

Strobel, G. A., and Hess, W. M. 1968. Biological activity of a phytotoxic
glycopeptide produced by Corynebacterium sepedom'cum. Plant Physiol.
43:1673-1688.

Thyr, B. D. 1968. Resistance to bacterial canker in tomato, and its evaluation.
Phytopathology 58:279-281 .

Thyr, B. D. 1969. Additional sources of resistance to bacterial canker of tomato
(Corynebacterium michiganense). Plant Dis. Rep. 53:234-237.

Thyr, B. D., Samuel, M. J., and Brown, P. G. 1975. New solanaceous host
records for Corynebacterium michiganense. Plant. Dis. Rep. 59:595-598.
Tsiantos, J. 1987. Transmission of bacterium Corynebacterium michiganense pv.
michiganense by seeds. J. Phytopathology 119:142-146.

van den Bulk, R. W., Jansen, J ., Lindhout, W. H., and Loffler, H. J. M. 1991.
Screening of tomato somaclones for resistance to bacterial canker (Clavibacter
michiganensis subsp. michiganensis). Plant Breed Z. Pflanzenzucht 107:190-196.
Versalovic, J., Schneider, M., De Bruijn, F. J ., and Lupski, J. R. 1994. Genomic
fingerprinting of bacteria using repetitive sequence-based polymerase chain
reaction. Meth. Mo]. and Cell. Biol. 5:25-40.

Versalovic, J ., T. Koeuth, A., and Lupski, J. R. 1991. Distribution of repetitive
DNA sequences in eubacteria and application to fingerprinting of bacterial
genomes. Nuc. Acids Res. 19:6823-6831.

Volcani, Z., Zutra, D., and Cohn, R. 1970. A new leaf and fruit spot disease of

pepper caused by Corynebacterium michiganense. Plant Dis. Rep. 54:804-806.

17

53.

54.

Wallis, F. M. 1977. Ultrastructural histopathology of tomato plants infected with
Corynebacterium michiganense. Physiol. Plant Pathol. 11:333-342.

Waters, C. M., and, and Bolkan, H. A. 1992. An improved semi selective
medium and method of extraction for detecting Clavibacter michiganensis subsp.
michiganensis in tomato seeds. (Abstr.) Phytopathology 82:1072.

Williams, C. C. 1928. The Grand Rapids disease of tomatoes. Ohio Agr. Expt.

Sta. Bull. 117237-38.

18

SECTION I

GENOMIC VARIATION OF CLA VIBAC T ER MICHIGANENSIS SUBSP.

MICHIGANENSIS IN MICHIGAN

19

ABSTRACT

Strains of Clavibacter michiganensis subsp. michiganensis were isolated from
infected foliage and fruit collected from 14 and 9 Michigan commercial tomato fields in
1997 and 1998, respectively, and subjected to rep-PCR fingerprinting. The cultivar
Mountain Spring dominated the plantings in southwest Michigan during 1997 and 1998
and the most commonly observed C. michiganensis subsp. michiganensis strain was type
C (302 out of 319). Avirulent type C strains (16%) were detected from two southwest
fields. All 85 strains from one northeast farm where several cultivars were sampled in
1997 and 1998 were of type D. In 1997, strains isolated from two fields of one north-
central grower were primarily type C (5 8) with 2 type D strains. Conversely, in 1998,
two north-central fields of the same grower contained mostly type D (35) and a few type
C strains (9). In 1997 and 1998, the tomato fields sampled in the southeast region of

Michigan contained 66 A, 63 B and 78 type C strains.

20

INTRODUCTION

In 1999, the Michigan tomato (Lycopersicon esculentum Mill.) industry generated
over 27 million dollars (1). Bacterial diseases, including bacterial canker of tomato
(causal agent: Clavibacter michiganensis subsp. michiganensis), are a yearly concern for
growers. Infection by this pathogen can result in plant wilting, stunting, reduced yields,
and plant death. Less severe disease symptoms include marginal leaf necrosis that may
be bordered by chlorosis, and fruit lesions appearing as superficial white spots (3-6 m)
that develop a necrotic center.

Field sources of inoculum include infected transplants, surface pathogen
populations on tomato and alternative weed hosts, and overwintered tomato debris (4).
The pathogen may be seed borne and can overwinter in fields of north-central United
States for up to 2 years (8). Clavibacter michiganensis subsp. michiganensis survives on
the surface and in the layers of the seed coat (2, 15) and can be transmitted to emerging
seedlings (2, 4, 16). Epiphytic bacteria thrive in the humid greenhouse (16) and may
spread to uninfected plants via overhead watering (Hausbeck, unpublished data). In the
Midwest, tomato transplants are grown in local greenhouses for use in establishing
production fields. Greenhouse transplants can appear healthy, yet harbor high bacterial
populations that may lead to severe disease symptoms in the field (10). Epiphytic
populations on field plants may spread by water splashing, pesticide spraying, pruning,
tying, sand blasting, and harvesting (9).

Bacterial genomes typically include a variety of low copy number repeated

sequences including interspersed repetitive DNA sequences (IRSs) that range from 15 to

21

several hundred base pairs (17). They are widely dispersed throughout prokaryotic
genomes, non-coding, and intercistronic; one example being interspersed, repetitive BOX
sequences (BOX) (1 7). The presence, widespread distribution and highly conserved
nature of these sequences suggest that these repetitive sequences are important to the
structure and evolution of the bacterial genome, yet their function(s) is not yet known

(1 7). Repetitive-sequence-based (rep)-PCR is a method used to identify and differentiate
microbes. The technique utilizes outwardly facing primers complementary to IRSs within
the target genome resulting in the amplification of differently sized DNA fragments
between the elements. When C. michiganensis subsp. michiganensis is subjected to rep-
PCR with BOX as a primer (BOX-PCR) and PCR products are electrophoretically size
fractionated in agarose gels the pathogen is distinguished into four fingerprint types (A,
B, C, D) (l 1).

Prior to this study, little was known concerning the genomic variation of the
pathogen in commercial fields. Characterization of the variation within a pathogen
population may be beneficial in addressing epidemiological questions. The objective of
this study was to characterize the variation in populations of C. michiganensis subsp.
michiganensis within and among Michigan commercial tomato fields using BOX-PCR
fingerprinting.

MATERIALS AND METHODS

Sample collection. Foliar and fruit samples were collected from 14 and 9

commercial fields from four regions of Michigan in 1997 and 1998, respectively (Table 1;

Figure 1). Samples were haphazardly collected according to cultivar or location in a

22

Table 1. Strains of Clavibacter michiganensis subsp. michiganensis isolated from fruit

and foliage collected from Michigan commercial fresh market and processing tomato
fields in 1997 and 1998.

 

Region/Grower/

 

 

 

Field/Date Collected PCR typea Cultivar Strain(s)"
Southwest (SW)
SW/l/l/97-Ju1y B Mountain 91 -2c
Spring
C Mountain 2-2‘, 8—2“, 11-2, 17-1‘, 27-1, 28-1,
Spring 47-2, 50-1, 63-1, 70-2, 74-2, 78-1,
83-1, 86-2, 98-2, 101-2, 103-1",
'104-1, 108-2, 109-2", 112-2, 119-
2‘, 130-1, 131-2, 136-2, 137-2°,
146-2, 152-2, 154-2c
SW1/1/97-August B Mountain 2-2, 17-2, 91-1
Spring
C Mountain 8-1, ll-2°, 27-2, 28-2, 47-2‘, 50-2,
Spring 74-1, 78-1, 83-1“, 86-1, 98-2", 101-
1, 103-1, 104-1, 108-1, 109-1,
112-1°,119-1,124-2,130-1°,13l-
2, 136-2, 137-2, 146-2, 152-1,
154-1c
SW/1/1/97-October C Mountain 2-1", 8-2, 11-2, 17-2“, 28-2, 47-1,
Spring 50-1‘, 63-1, 74-1, 78-1, 83-2, 86-
2“, 91-1, 98-2, 103-1, 104-2, 108-
2d,109-1,ll9-1°,124-2,130-1,
131-2. 136-2. 137-1‘, 146-2, 152-
2, 154-1
SW/ 1 /2/ 97- B Mountain m4-2
September Spring
C Mountain el-l , e2-2“, e3-2, e4-l, e6-2, e9-1°,
Spring elO-l‘, dl-2, d2-2°, d4-1‘, nd1-1°,

23

nd2-2, nd3-2‘, nd4-1, m1-2, m2-1,
m3-ld, m5-1°, m7-2°, m8-2, m10-

2, fl-2, f3-1d, f4-l, fS-Z, f7-1c, f9-
2, flO-l, H2-4

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR type‘I Cultivar Strain(s)b
SW/1/3/97- C Mountain e2-3, e2-4‘, e3-l, e3-2, e3-3, e4-1,
September Spring e4-2, e5-1, e5-2, e5-3°, fl-4, 12-2,
f3-5, f4-1, f4-3, f4-4-1, f4-5, f5-3,
m1-2, m1-3, ml-4-l, ml-4-2,
m1-5-3, m3-3, m3-4, m4-2, m4-3,
m5-1, m5-2, mS-3°
SW/2/1/97- C Mountain el-lc, e2-1, e2-2, e3-1, e4-l, e4-2,
September Spring eS-l, e5-2“, e6-2, e7-2, e8-l, e8-2,
e9-l, e9-2, e10-2, m1-2‘, m2-1,
m2-2, m3-l, m4-1, m4-2, m5-2,
m5-3s, m6—2, m7-2°, m8-2, m9-
1,m9-2, m10-2d
SW/2/2/97- B Mountain m4, m6-2, N08
September Spring
C Mountain OL1-3, 0L3, 0L4, OL5-l, 0L6-
Spring 1, 0L7, OL8-2, 0L9, OL10-2, el-
2, e2-2, e3-1, e4-2, m2-1, m3-1,
m5-2, m7-2, m8-1, m9, NOl-l,
N03, N04, N05, NO6-2, N07,
N09, N010
SW/3/1/97- B roma type r2-2
September
C Mountain f1-l,f1-,2 12- 1, f3-l, f4-3, f5-1,
Spring f5- 2, f8- 1, f9-1,m1-1, ml-2, m2-
2, m3-2, m4-2, m5- 1, m6—l, m6-
2, m7-1, m8-,1 m9-,1 m9- 2, m10-
1
roma type rl-l, r3-2°, r4-2, r4-3, r5-1, r7-1,
r8-1c
SW/4/1/97- A mixed fresh NE2-2, NWI-ld, NW2-2, NW1g-
September market 1, NW3 g-lc
roma type Rw2-2, Rw8-l, Rw8-3
C mixed fresh NE1-2, NE2-2, NW1-2
market

24

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR typea Cultivar Strain(s)"
roma type Rwl-l, Rw2-l, Rw5-l, Rw7-1,
Rw9-1, Rw10—2, Rel-1, Rel-2,
Re7-l, leg-2, Rm4g-3
Stokes fresh SEl-l, SE1-2
market
Mountain SWI-l, SW1-2, SW2-2, SW3-2
Spfing
SW/1/2/98-August C Mountain BLl, BL2°, BCl, BC2, BRl,
Spring BR2, CLl, CL2, CR1, CR2“, f1-
l°, t2-2°, f3-2°, f4-l, f5-1d, f6-1,
ml-l, m2-2‘, m4-l, m5-1, n3-1°,
n5-1°, r2-2, r3-2", r4-ld, r5-2d
SW/2/3/98-August C Mountain fl-l", f2-1d, f3-2, f4-2d, f5-1, f6-
Spring 2", f7-l, f8-1, f9-2, f10-l, ml-l,
m2-l“, m3-2, m5-1, m6-1d, m7-l,
m8-2, m9-1, m10-2, rl-l, r2-2",
r3-2, r4-2d, r5-1, r6-ld, r8-2, r9-2,
rlO-l
Northeast (NE)
NE/1/1/97-September D mixed fresh Aeh2, Aeh3, Aeh4, Aeh5, Aetl
market
Aztec AZl , AZ4, A25
mixed fresh Behl, Beh2, Beh3, Betl-S-l,
market Bet3-5-1
Flavorrnore FM1-4-1, FM2
Mountain MF1-5-1
Fresh
pear type P1
roma type R1-5-2, R3-5-l
Red Cherry RC2-5-1, RC3-4-1
Red Rider RRl-S-l, RR3-5-l

25

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR type' Cultivar Strain(s)"
Springfield SP1
Asgrow 145 145-1, 145-2
NE/l/1/98-June D mixed fresh GHl f“, GH2-4c, GH3-5c
market 9
Majesty Ml-l,Ml-4
Mountain MF4-2c
Fresh
Mountain MS6-2
Spring
Mountain MSu4-l°, MSu5-2
Supreme
Red Rider 4d-2d, 5a-1, 5c-2
roma type R1 -2", R4-3
Springfield SP2-l
Super Beef Sbfl -2, Sbf6-l‘, SBf6-5
Sunbrite Sbrtl-l, SB2-l°
Sun Gem SG2-l“, SG3-ld
Sun Start SStlf4", SSt2fl h, SSS-2b
Trust T1-2", T1-3, T2f1
NE/1/1/98-July D Beef King BK2-1, BK5-2c
heirloom Hl-l , H5-2
type
Majesty M5-2
Mountain MF5-2
Fresh
Mountain MSl-1°, MS3-2
Spring

26

Table 1. continued

 

 

27

Region/Grower/
Field/Date Collected PCR type‘ Cultivar Strain(s)"
Mountain MSUl-l, MSU3-l
Supreme
Plum Dandy R2-1, R4-1°
Red Rider RR1-2, RR4-2"
Springfield SP4-2c
Sunbrite SBR2-1d, Sbrt2-2°, SBR5-1°,
Sbrt5f°h
Sun Gem SG2-l, SG2-2d, SG3-1, SG4-l
Sunstart SS4-1°, 5-2-1‘, 6—4—1c
Super Beef SBFl-l, SBF4-l
Trust T2-l°, T3-1, T4-2
North-Central (NC)
NC/1/1/97-August C Jet Star 44-2, 45-1, 46-2, 48-1, 52-2, 56-
2, 58-1, 59-1, 60-1
roma type 26-2, 27-2, 30-2, 33-2, 34-1, 35-
2°, 37-1, 39-1, 40-1c
Sun Gem 2-2, 6-2, 9-2, 10-2, 14-2, 15-2,
16-1, 18-2, 20-2, 22-2
Jet Star 53-1
roma type 36-1
NC/1/2/97-August Jet Star l-l‘, 2-1, 5-1, 6-2, 8-2‘, 9-1, 10-2,
11-1, 12-2°,14-2, 18-1,19-2c
Sun Start or 21-1,25-1°,3l-l,32-1,35-2°,38-
Red Rider 1, 41-1, 43-2, 44-2, 46—2, 48-2,

49-1, 50-1, 56-2, 57-1, 58-1, 59-
2, 60-1

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR type“ Cultivar Strain(s)"
NC/ l/3/98-September C Mountain MSu2-2, MSulO-2
Supreme
Sun Start SStl-l, SSt2-2, SSt3-l, SSt4,
SSt6-2, SSt7-l, SSt9-1
D Mountain MSu1-2, MSu3-1, MSu4-2,
Supreme MSu5-l, MSu6, MSu7-2, MSu8,
MSu9-2
Red Rider RRl-l, RR2-l, RR3-l, RR4,
RR5-2, RR6-1, RR7, RR8-2,
RR9-2, RRlO-l
Sun Start SSt5-l, SSt8-2, SSth-l
NC/1/4/98-September D Roma Rl-l, R1-2, R2-1°, R3-2°, R4-1,
R4-2, R5-l°, R6-1°, R7-1, R7-2,
R8-2°, R9-1, R9-2, RIO-1°
Southeast (SE)
SE/l/1/97-September A orange Echl-2b, Echl-3, Ech2-l", Ech2-3
cherry type
red cherry Erl -3
type
“Spring” cv. Esple-l“, Esp2-2, Esp2-3, Esp3-2
yellow plum EYp12-2, EYpl3-2
type
B red cherry Erl -2
type
“Spring” Esp4-l, Esp2e-3
type
yellow plum EYpl 1 -1
type
C “Spring” cv. Espl-l, Esp2e-2“, Esp3e-l‘,

28

Esp3e-3

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR type' Cultivar Strain(s)"
SE/ 1/2/97-September A Cherry Pink CPI-2, CP2-2d, CP3-2, CP4-2,
CPS-2
Golden GN1-2, GN2-1b, GN3-1, GN4-2°,
Nugget GNS-l
B yellow pear YPl-lc
type
C yellow plum YPll-l, YP12-1
type
SE/2/ l/97-September A Sunrise BbA4-1
B Doubletake Dl-2, D2-l°, D4-2
Mountain C2-2°, C5-1, C6, CbB5-1
Spring
Nickelow El-l, E4-l, E7-1c
Sunrise B4-1, B8-2°, BbA2-l‘, BbC5-2,
BbC6f°"
Sunstart A4f‘, A5
C Doubletake D3-1
Sunrise B7-1, BbAl-l, BbC2-1, Bble‘,
BbC4-1c
Sunstart Al-2, A3f‘, AbBlf“, AbB5(1)6f'
SE/3/1/97-September A Mountain D3-2
Fresh
Sunrise Al.1-l,A3.l-1,A4.1-2
B Mountain Dl-l, D2.1-1
Fresh
Sunrise C3.2-2
C Mountain D4—2, D5-l
Fresh
Springfiesh Bl.1-l, B4-2

29

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR type' Cultivar Strain(s)"
Sunrise A2.l-2,Cl.2-1°,Cll(l)lf‘
SE/1/2/98-August A Asgrow A9-2, AlO-2
6000
Golden GN2-l , GN4-2c
Nugget
Italian Gold IG4-1
Market MP2-l
Pnde
Mountain MF3-2, MF6-1
Fresh
Sun Gold SG2-2°, SG3-l
yellow pear YP1-2, YP2-1c
type
B Asgrow A4-2
6000
Enterprise E2-2d
heirloom Hl-l , H2-2°, H3-l
type
Italian Gold IG2-2c. IG3-1d
Market MP1-2, MP8-2
Pnde
Mountain MF1-2, MF2-1, MF4-l
Fresh
Mountain MGl-l , MG2-2d
Gold
roma type R2-2c
Sun Gold SGl-ld
SE/ 1/3/98-August A Red Rider RRl, RR3, RR6, RRlO
Sun Brite SB(GA)2°, SB(GA)4-2c

30

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR type‘ Cultivar Strain(s)b
Sun Gem SGl, SG(GA)2", SG3-l,
SG(GA)4-2°, SG(GA)7°, SGIO,
SG(GA)10-1°°
Sun Start 88], SS(GA)l-2°°, SSZ, SS4,
SS(GA)6°, SSS
B Sun Gem SG(GA)6-2c
Sun Start SS(GA)5°
C Red Rider RR2
Sun Brite SB(GA)7c
Sun Gem SG2-2, SG4, SG(GA)5°,
SG(GA)8‘, sow/s.»dc
Sun Start SS3
SE/4/l/98-August A Asgrow 3r A3-9-lh
Glamour G7-1", GlO-lh
B Asgrow-2f A2-2", A2-6-l, A2-10-2
Asgrow-3f A3-2", A3-7-lh
Glamour G8-1
Golden GJ 10h
Jubilee
Mountain MFlh, MF3", MF8-l
Fresh
Mountain M33", M84”, M38h
Spring
Sunbrite SBlh, SB6-2
C Asgrow-2f A2-3h, A2-4h, A2-6-2, A2-10-l
Asgrow-3f A3-1h, A3-4-2, A3-6-l, A3-6-2,
A3-9-2
Glamour GL1“, GL2”, GL3h, GL4”, GL7-2,

31

GL8-2, GL10-2

Table 1. continued

 

 

Region/Grower/
Field/Date Collected PCR type' Cultivar Strain(s)b

Golden GJ7h

Jubilee

Mountain MF4h, M176“, MF9-2, MF 1 0-1

Fresh

Mountain MS8-2, MS9h

Spfing

Springfield SP7-2, SP8“, SP9h

Sunbrite S33”, SBS”, SB6-lh

Sun Gemc A1-6“, Al-7", Al-IOh
SE/5/1/98-August A $012 I-1,I-2,I-5-1h

401 III-2

9701 IV-3, IV-6- l h

B 696 II-l , II-5
401 III-5h
9701 IV- 1 , IV-Z, IV-7-2h
C 8012 I-3h, I-4, I-5-2h, I-6h, I-8h, I-9h

696 II- 1 , II-Z, II-4, II-7, II-9h

401 III-3, III-7

9701 IV-Sh, IV-7-l h, IV-8”, IV-1 Oh

 

’ Polymerase chain reaction designated type using BOX primer (Louws et al., 1998).

b Strains were named with respect to a particular cultivar or location in a field followed by
a dash and the colony number the strain was isolated from; i.e. cultivar Mountain Spring-
colony 2 = MS-2.

cBOX-PCR fingerprints identical in duplicate PCR reactions.

dBOX-PCR fingerprints of same PCR type yet differ in polymorphic bands when
replicated.

° Transplants grown in Georgia.

f Transplants grown in Florida.

3 Strains isolated from greenhouse grown plants.

" Strains isolated from a fruit lesion; all other strains were isolated fi'om foliage.

32

 

t1

 

 

 

 

 

 

_I
Ll...

 

 

 

 

 

 

 

 

 

l Northeast

 

 

 

 

 

 

 

North-Central 1

Southwest fl

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Southeast

 

 

Figure 1. Four regions of Michigan from which Clavibacter michiganensis subp.
michiganensis strains were isolated from tomato plants and fruit collected from
processing and fresh market commercial fields in 1997 and 1998.

33

field. One exception was field SWl/ 1/97, which was sampled July, August and October,
1997 along diagonal transects. In this field only, plants to be sampled were labeled and
sampled each month. Foliar samples consisted of 2 to 3 basal leaves exhibiting bacterial
canker foliar symptoms. Fruit exhibiting bird’s eye spots were also collected. All
samples were placed in plastic bags, transported in ice chests and stored at 4°C for a
maximum of 3 days.

Bacterial isolation from foliage and fruit. Stems and leaves were chopped and
homogenized for approximately 60 seconds in a Lab-Blender 400 stomacher (Tekmar
Co., Cincinnati, OK) using phosphate buffer (0.05 M, ph 7.4; 2 ml/g of plant tissue)
amended with Tween-20 detergent (0.02%). Each sample was either undiluted or
subjected to a lO-fold serial dilution and grown on nutrient broth yeast extract (N BY) (2)
agar modified by omitting glucose (MNBY) and/or semi-selective media for C.
michiganensis subsp. michiganensis (SCM) (5) both amended with nalidixic acid
(0.03 g/L) and cycloheximide (0.1g/L). One ml of the plant extract was vortexed with 0.5
ml of 40% glycerol and stored at -20°C. After a pure culture was obtained, each isolate
was handled in one of two methods: 1) inoculated into 5 mL of MNBY broth and placed
on a rotary shaker at 175 rpm for 48 hours, after which 1 ml was stored in 0.5 mL 40%
glycerol; 2) inoculated into a sterile cryo-storage solution containing distilled water (19%
by volume), tryptic soy broth (TSB; 78% by volume) and glycerol (3% by volume). The
isolates were then stored at -20°C for later PCR.

Fruit lesions resembling bird’s eye spots were probed using sterilized

toothpicks directly under the surface of the lesion. The toothpicks were streaked on

34

MNBY and colonies resembling C. michiganensis subsp. michiganensis were stored as
described above.

Whole-cell PCR amplification and electrophoresis. Strains were isolated from
approximately 30 individual plant and/or fruit samples from each field and analyzed
using the whole-cell method of rep-polymerase chain reaction (rep-PCR) as described by
Louws et al. with the following modifications (1 l). Colonies grown for PCR
amplification were grown on either MN BY. or SCM agar. Amplification was performed
in a thermal cycler (Genemate-Techne, Princeton, NJ) equipped with a heated lid,
therefore, eliminating the need for mineral oil. Each 25 :1 reaction included DNA
AmpliTaq polymerase (0.4:1; Perkin Elmer, Norwalk, CT). The primer, BOXA1R(5'-
CTACGGCAAGGCGACGCTGACG-3') (12), was synthesized at the Macromolecular
Facility, Department of Biochemistry, Michigan State University, East Lansing. DNA
fragments within the PCR products (6:1) were separated at 4°C on 1.5% agarose gels in
0.5x Tris-acetate-EDTA (TAE; 0.4 M Tris-acetate and 1 mM EDTA, ph8.0) buffer at 83
volts for 13 hours. The gels were stained with etlridium bromide (0.5 :g/ml), soaked in 5x
TAE and photographed under UV illumination using Polaroid Type 55 or 57 film.
Genomic DNA isolated from strain 936 (1 1) of PCR type A was included as a positive
control with each PCR run. A negative control consisting of the reaction mixture with
the omission of bacterial cells was also included with each run. Gel photographs were
analyzed by eye to determine the designation of PCR fingerprint types with the
corresponding strains. Twenty—one percent of the total strains were amplified twice using

different reaction mixtures and performed on different days (Table 1). When the

35

fingerprints of the replicated strains were compared, all of the isolates were the same PCR
fingerprint type, yet 24% of the replicated strains differed in some polymorphic bands
(Table 1).

Hypersensitivity response (HR) assays. Four o’clock (Mirabilisjalapa) foliage
has been shown to develop hypersensitive reaction lesions when inoculated with C.
michiganensis subsp. michiganensis (6) and is a method used as a standard pathogenicity
test for the bacterial canker pathogen in Georgia and other seed-testing facilities (9).
Broth cultures of C. michiganensis subsp. michiganensis were grown in liquid MNBY
broth for 48 hours with shaking at room temperature. Cells were harvested by spinning
on high in a micro-centrifuge for 3 minutes and washed two times with sterile distilled
water. The suspensions were adjusted with sterile distilled water to 50% transmittance at
a wavelength of 600 nm measured with a spectrometer 20 and kept on ice. The bacterial
concentration of suspensions prepared in this way contain approximately 108 CFU/ml (6).
Four o’clock plants were grown in clay pots in a commercial potting mixture in a research
greenhouse. A needleless hypodermic syringe was used to infiltrate the leaves with the
prepared bacterial suspensions. Sterile water and a known HR-positive strain (strain 68)
(11) were infiltrated on a leaf of each plant as negative and positive controls, respectively.
The presence or absence of local lesions were observed and recorded two days after
infiltration. All tests were performed in duplicate on separate plants.

RESULTS
The cultivar Mountain Spring dominated commercial plantings in southwest

Michigan during 1997 and 1998 and the most commonly observed C. michiganensis

36

subsp. michiganensis strain was type C (302 out of 319 strains) (Table 2). In 1997, the
southwest fields containing only ‘Mountain Spring’ tomatoes (SW/ 1/ l , SW/ 1/2, SW/2/l)
or ‘Mountain Spring’ and roma-type tomatoes (SW/3/ 1) harbored C. michiganensis
subsp. michiganensis strains representing primarily the type C strains (27 to 30
strains/field) and few type B strains (0 to 3 strains/field) (Table 2). Field SW/4/1/97
contained at least four cultivars and isolates were mostly type C strains (20) and some
type A strains (8) while two plants had a mix of type A and C strains (Table 2). In 1998,
two fields (SW/1/2, SW/2/3) each containing only ‘Mountain Spring’ harbored only type
C strains (Table 2).

When ‘Mountain Spring’ tomatoes were planted in the same fumigated field in
consecutive years (SW/1/2/97-98), the C fingerprint type was dominant in 1997 (29
strains) and 1998 (26 strains) with one type B strain found only in 1997 (Table 3). When
a commercial ‘Mountain Spring’ tomato field (SW/1/1/97) was sampled three times over
the course of a growing season, the C. michiganensis subsp. michiganensis PCR types
remained relatively stable consisting primarily of type C strains (26 to 29 strains/sample
time) and few type B strains (0 to 3 strains/sample time) (Table 4). Three of the 31 plants
sampled harbored both type B and C strains over the time sequence (Table 4).

At the beginning of the1997 growing season (July) 8 virulent and 6 avirulent C.
michiganensis subsp. michiganensis strains were detected in field SW/1/1/97 (data not
shown). During mid- to late season (August) virulent strains increased nearly three-fold

(21) while avirulent strains were decreased by half (3). By the end of the season

37

Table 2. BOX-PCR fingerprint types of Clavibacter michiganensis subsp. michiganensis
and isolated from fruit and foliage collected from Michigan commercial fresh market and
processing tomato fields in 1997 and 1998 according to region.

 

 

 

 

 

 

 

Region/Grower/ If)? PCR fingerprint types
Field/Date Collected Cultivars A B C D Totals
Southwest (SW)
SW/1/1/97-July 1 0 l 29 0 30
SW/l/1/97-August l 0 3 26 0 29
SW/ 1/ 1 /97-October 1 0 0 29 0 29
SW/1/2/97-September l 0 1 29 0 30
SW/1/3/97-September l 0 0 30 0 30
SW/2/1/97-September 1 O 0 29 0 29
SW/2/2/97-September l 0 3 27 0 30
SW/3/1/97-September 2 0 1 29 0 30
SW/4/1/97-September 4 8 0 20 0 28
SW/1/2/98-August l 0 0 26 0 26
SW/ 1/3/98-August l 0 0 28 0 28
in t r 9 3=0.;__2.__.u2_
Northeast 1 NE)
NE/ 1 /1 /97-September 9 0 0 0 26 26
NE/1/1/98-June 12 0 O 0 28a 28"
NE/1/1/98-July 14 0 0 O 31 31
Region Total 0 0 0 85 85

 

38

Table 2. continued

 

 

 

 

 

 

NO- PCR fingerprint types
Region/Grower/ of
Field/Date Collected Cultivars A B C D Totals
North-central (N C)
NC/l /1 /97-August 3 0 0 28 2 30
NC/ 1/2/97-August 3 0 0 3 0 0 30
NC/1/3/98-September 3 0 0 9 21 30
NC/l /4/98-September l 0 0 0 l4 l4
Eggign Total 0 0 67 37 l 04
Southeast 1 SE)
SE/l/1/97-September 4 l l 4 4 0 l9
SE/1/2/97-September 4 10 l 0 13
SE/2/1/97-September 5 1 16 10 0 27
SE/3/ l /97-September 4 4 3 7 0 14
SE/l /2/98-August 1 l 12 16 0 0 28
SE/ 1 /3/98-August 4 l9 2 0 29
SE/4/ l /98-August 9 3 15 33 0 51
SE/5/1/98-August 4 6 6 l7 0 29
Region Total 66 63 81 0 210

 

‘ Include three type D strains isolated from mature greenhouse grown tomato plants.

39

Table 3. BOX-PCR fingerprint types of Clavibacter michiganensis subsp. michiganensis
strains isolated from foliage collected in consecutive years (1997 and 1998) from
Michigan commercial fresh market tomato fields.

 

NO- BOX-PCR Fugerprint Types

 

 

 

 

Region/Grower/ of

Field/Date Collected Cultivars A B C D Totals
SW/1/2/97-September 1 O l 29 0 30
SW/ l/2/98-September l 0 0 26 0 26
NE/ 1/ l /97-September 9 0 0 0 26 26
NE/ 1/1/98-June 12 0 0 0 28“ 28
NE/1/1/98-July l4 0 0 0 31 31
SE/1/2/97-September 4 1 0 l 2 1 3
SE/ 1 /2/98-September 1 1 12 16 0 28

 

" Includes three type D strains isolated from mature greenhouse grown tomato plants.

40

“33538 H 3m .

 

 

 

u u o o u o o u o u U o o 5880
o o o o o o o o o o o o o o o saw?
u u u U u o o o u U o u U 0 b2
2 an a an E 3 mu 3 an 2 a 8 a 2 2 5.82

oEEam

«PC. Evibwfih mus—*6: wficgmmoteu cc: eon—.52 29:5 :35 .3236:—

 

 

 

 

 

 

 

U U U U U U U U U U U U U U 89260

U m U U U U U U U U m U U m “wows/x
m o o o o o o o o o o o o o o be

m— I m — n— 2 c— a w h o m e n n — 5.82

«95. BER—ogr— momkcm wficgamotcu ES .638: Z fiqfiam =3..— _a=e_>==: «EEam

mm o mm o o LBOSOSQ C :26

am o mm m o namieaczam

2 o SN _ o 23-8255
233. G U m < 123.30 Sag—«E

totemviemwom

 

men—.3. .=_.a._ow=_,m mUm-XOm me ban—Eam

 

.33 5 Eva 0388 .wctqm E3552. 30.5888 59:32 a 89a 38: ooh:
8 @2828 owmzfi 80¢ team—0mm mafia mmermaomEoNE dmnsm amaamaoMEoE $33330 («o 396 “Sundown—M munznom .v oEah

41

(October) 13 virulent and 6 avirulent strains were detected. In another field (SW/ 1/2/97),
a greater number of virulent strains (1 7) than avirulent strains (2) were detected. Similar
results were found when this field was sampled the following year, resulting in the
detection of 18 virulent and l avirulent strain(s).

All strains from the one northeast farm (N E/ 1/ l) where several cultivars (9 to 14)
were sampled in 1997 and 1998 were of type D (Table 3). In 1998, some fruit produced
in a greenhouse on this grower’s farm had symptoms of bird’s eye spots from which C.
michiganensis subsp. michiganensis type D strains were isolated (Table 1; strains GHl f,
GH2-4, GH3-5).

In 1997, the strains isolated from two fields (NC/ 1/1 , NC/1/2) of one grower in
the north-central region were primarily type C strains (28 and 30 strains, respectively)
with few type D strains (2 and 0 strains, respectively) (Table 2). Conversely, in 1998,
fields NC/ 1/3 and NC/ 1/4 contained mostly type D strains (21 and 14, respectively) and
few type C strains (9 and 0, respectively) (Table 2).

The plants sampled from fields in southeast Michigan contained 66 type A , 63
type B and 78 type C strains (Table 2). In field SE/1/2 the type A fingerprint was
dominant (10 out of 13) with type B and C strains detected in low numbers (1 and 2,
respectively) in 1997 (Table 3). When the same field was sampled the following year,
type A strains (12) and an increased number of type B strains (16) were detected (Table
3). Two plants in field SE/1/1/97 harbored mixed PCR type strains; one with A and B and
another with B and C. Field SE/3/ 1/97 contained two plants with mixed fingerprint types

of A with C and B with C, respectively. In a third field (SE/5/1/98), isolated strains

42

obtained from eight plants were identified as mixed type strains; three with A and C and
the other five with B and C, respectively.
DISCUSSION

When recommendations to control bacterial canker on tomato are considered, a
major focus has been on the source of inoculum. It has been well documented that the
pathogen can be seed borne (3, 7) or transmitted via infested tomato debris overwintered
from a previous planting (8). Historically, Michigan growers utilized transplants that had
been grown in southern outdoor seedbeds. After a severe bacterial canker outbreak in the
mid-westem United States and Canada was traced back to infected transplants from
Georgia (7), infected transplants were recognized as an additional source of inoculum.
Although Michigan growers now rely on local greenhouse-grown tomato transplants to
avoid bacterial canker outbreaks, this change has not eliminated the disease. The warm
and humid environment of the greenhouse in addition to frequent overhead watering
creates favorable growing conditions for bacterial populations, therefore, infected
transplants continue to be a concern as a source of inoculum in Michigan. Inferences may
be made between some of the bacterial canker outbreaks observed in this study and the
three sources of inoculum mentioned above.

Southwest Michigan is a major fresh market growing area and is unique because
of the dominance of ‘Mountain Spring.’ This cultivar was present in every field sampled
and was the only cultivar in six of the eight fields included in our study. The dominant
(95%) fingerprint type found in this region was type C. In a C. michiganensis subsp.

michiganensis collection spanning from 1981 to 1994 (11), type C strains (11) detected in

43

southwest Michigan fields were present at a higher incidence than type A (5) and B (3)
strains. In our study, A and B strains comprised 5% of our isolates from this region. In
one southwest field (SW/1/2) that was sampled in 1997 and 1998, symptom severity
varied between years. This field had been called to our attention in 1997 because of
young, newly established seedlings that were exhibiting severe wilting and eventually
died. It is our experience in Michigan that such severe symptoms are the result of high C.
michiganensis subsp. michiganensis populations on the seedling at the time of
transplanting (10). Whether the C. michiganensis subsp. michiganensis originated from a
contaminated greenhouse or from infected seed is not known. However, in 1998, disease
symptoms in that same field were minor, consisting of marginal leaf necrosis on an
insignificant number of plants. This grower compensates for a shortened of lack of
roatation by fumigating the plant beds yearly. Although overwintered host debris was not
readily observed, the soil between the plant beds was not fumigated leaving a possible
source of inoculum. It is also possible that seedlings with populations far below the
threshold required for more severe symptoms to occur were used to establish the field in
1998.

While northeast Michigan has only one significant tomato grower, isolates
collected from this farm offer a unique perspective on the potential role of greenhouse
and field as potential sources of C. michiganensis subsp. michiganensis inoculum.
Although multiple cultivars (314) were grown, only one fingerprint type D was observed
among the 85 isolates collected during 1997 and 1998. The grower in the northeast

region produced tomatoes within a 5 acre plot that was rotated within a 20 acre field and

44

did not fumigate. Transplants used in the production fields were grown in greenhouses
located on the farm but not adjacent to the production field. After the transplants were
moved to the field, the grower used the greenhouse to produce full-sized plants and fruit
from which C. michiganensis subsp. michiganensis strains were isolated in 1998. The
greenhouse-grown plants did not Show foliar symptoms, yet did Show minimal fruit
spotting. A recent study reported that plants are most susceptible to fruit at the flowering
stage and small fruit stage of plant development (13). This suggests that the pathogen
populations were high enough at the time of flowering to cause minimal fruit spotting, yet
not high enough to cause severe symptoms on the mature plants. Sanitation practices in
this greenhouse had not been closely adhered to. The routine of extended greenhouse
production together with the lack of field rotation create a situation in which bacterial
populations may have overfallowed on buried debris and/or alternative weed hosts in the
field and/or the greenhouse, thereby providing a reservoir of C. michiganensis subsp.
michiganensis from 1997 to 1998.

Type D strains were not found in the major fresh market tomato growing regions
of the southwest or southeast but were found on another Michigan farm located in the
north-central region approximately 80 miles away from the northeast grower. The grower
from the north-central region purchased transplants from the grower in the northeast
region, although the year of the original connection between these growers is not clear.
The north-central grower practiced short rotation (>2 years), did not fumigate the plant
beds and typically grew multiple cultivars (s 3). Infected transplants that were grown in

greenhouses of the northeast grower likely harbored type D strains that were then

45

transmitted to north-central field. Louws et al. (1 1) identified three type D strains from
this farm from samples collected in 1994. The north-central grower did not have type D
strains exclusively as did the northeast grower, but also had type C strains. A shift in the
incidence of the type D strains was observed from 1997 (3%) to 1998 (80%). Without
knowing the number of seedlings and cultivars provided by the northeast grower, it is
difficult to form a hypothesis regarding the reason for this change.

Another primer, (GTG)5 has been shown to distinguish BOX-PCR type D into
three groups when subjected to rep-PCR (Medina-Mora, thesis). Further analysis of the
type D strains from the north-central and northeast fields using this primer may provide
additional information on the variation between the type D strains found on these farms.

Typical cultural practices of the southeast region included long rotation (>3
years), no fumigation and a diversity of cultivars within and between farms. Unique to
any other field in this study, field SE/4/1/98, containing 3 type A, 15 type B and 33 type
C strains, had never been planted to tomatoes and was the first year this grower had ever
grown this crop (Table 2). With the exception of three F lorida-grown cultivars, all of the
seedlings used to establish this field were grown in the grower’s greenhouses that
previously had been used only for production of flowering ornamentals and melon
transplants. With the elimination of overwintered debris and greenhouse transplants as
likely inoculum sources, seed remains the most likely source of inoclum for the disease
outbreak in this field. Disease symptoms were severe within some cultivars (wilting and
stunting), yet minor within others (marginal necrosis). The variation of symptom severity

between cultivars strengthens the possibility of seed contamination. Seed from one or

46

more cultivars that were contaminated would likely exhibit severe symptoms in the field
as pathogen populations had time enough to reach threshold levels. Spread from infected
plants to uninfected plants in the greenhouse and/or in the field is suggested by the
presence of bird’s eye fruit spotting on all cultivars. Ricker and Reidel (14) determined
that secondary spread causes less severe symptoms of marginal necrosis and fruit
spotting, yet does not lead to yield loss.

Type A strains have been described as being most frequently associated with
processing cultivars (11) and in our study were only found in one southwestern Michigan
field (SW/4/ 1/97) that included a roma-type cultivar considered to be a processing type.
However, in this field, the type A strains were detected on both the roma-type and the
fresh market cultivars. Louws et al. (11) included four C. michiganensis subsp.
michiganensis strains obtained in 1994 from a southwest Michigan field (SW/4) that
contained roma-type tomatoes and also detected a type A strain: the only type A strain
detected in commercial fields in that region that year.

Louws, et al. (11) found a significant number (30%) of avirulent strains associated
with the type A strain. In our study 16% of the type C strains were avirulent, while type
C strains were found by Louws et al. (1 l) to be predominantly virulent. This study is the
first to report a substantial number of avirulent type C strains in field populations.

Although 24% of the strains replicated in two BOX-PCR runs exhibited some
differing polymorphic bands, some polymorphisms have been shown to be stable
(Fulbright, personal communication). These stable polymorphisms have been used in

epidemiological studies as a means of tracing a particular strain throughout the

47

experiments.

Prior to the technology of BOX-PCR fingerprinting, it would have been
impossible to make inferences concerning the transmission of bacterial strains from the
three type of inoculum sources discussed above. Further research is needed to test the
theories stated in this study. Any biological significance of the four fingerprint types is
not known and it is of interest to determine if one fingerprint type has a biological
advantage over another. This type of information may help to explain the presence of

dominant fingerprint types in some of the fields examined in this study.

48

LITERATURE CITED
Bay, D. M., and Wyant, D. (eds). 1999. Michigan Agricultural Statistics. MDA,
Michigan pp. 42-45.
Bryan, M. K. 1930. Studies on bacterial canker of tomato. J. Agr. 41 1825-851.
Chang, R. J ., Ries, S. M., and Pataky, J. K. 1991. Dissemination of Clavibacter
michiganensis subsp. michiganensis by practices used to produce tomato
transplants. Phytopathology 81 :1276-1281 .
Chang, R. J ., Ries, S. M., and Pataky, J. K. 1992. Local sources of Clavibacter
michiganesis subsp. michiganensis in the development of bacterial canker on
tomatoes. Phytopathology 82:553-560.
Fatmi, M., and Schaad, N. W. 1988. Semiselective agar medium for isolation of
Clavibacter michiganense subsp. michiganense from tomato seed.
Phytopathology 78:121-126.
Gitaitis, R. D. 1990. Induction of a hypersensitivelike reaction in four-o'clock by
Clavibacter michiganensis subsp. michiganensis. Plant Dis. 74:58-60.
Gitaitis, R. D., Beaver, R. W., and Voloudakis, A. E. 1991. Detection of
C lavibacter michiganensis subsp. michiganensis in symptomless tomato
transplants. Plant Dis. 75:834-838.
Gleason, M. L., Braun, E. J ., Carlton, W. M., and Peterson, R. H. 1991. Survival
and dissemination of Clavibacter michiganensis subsp. michiganensis in

tomatoes. Phytopathology 81 : l 5 19-1 523.

49

10.

ll.

12.

13.

14.

Gleason, M. L., Gitaitis, R. D., and Ricker, M. D. 1993. Recent progress in
understanding and controlling bacterial canker of tomato in eastern North
America. Plant Dis. 77:1069-1076.

Hausbeck, M. K., Bell, J ., Medina-Mora, C., Podolsky, R., and Fulbright, D. W.
2000. Effect of bactericides on population sizes and spread of Clavibacter
michiganensis subsp. michiganenesis on tomatoes in the greenhouse and on
disease development and crop yields in the field. Phytopathology 90:38-44.
Louws, F. J ., Bell, J ., Medina Mora, C. M., Smart, C. D., Opgenorth, D.,
Ishimaru, C. A., Hausbeck, M. K., deBruijn, F. J ., and Fulbright, D. W. 1998.
rep-PCR-mediated genomic fingerprinting: a rapid and effective method to
identify Clavibacter michiganensis. Phytopathology 88:862-868.

Martin, B., Humbert, 0., Camara, M., Guenzi, E., Walker, J ., Mitchell, T.,
Andrew, P., Prudhomme, M., Alloing, G., Hakenbeck, R., Morrison, D. A.,
Bounois G. J ., Glaverys, J-P. 1992. A highly conserved repeated DNA element
located in the chromosome of Streptococcus pneumoniae. Nucleic Acids Res.
20:3479-3483.

Medina-Mora, C. M., Hausbeck, M. K., and Fulbright, D. W. 2000. Bird's eye
lesions of tomato fi'uit produced by aerosol and direct application of Clavibacter
michiganensis subsp. michiganensis. Plant Dis. (In press).

Ricker, M. D., and Riedel, R. M. 1993. Effect of secondary spread of
Clavibacter michiganensis subsp. michiganensis on yield of northern processing

tomatoes. Plant Dis. 77:364-366.

50

15.

16.

17.

Strider, D. L. 1969. Bacterial canker of tomato caused by Corynebacterium
michiganense; a literature review and bibliography. North Carolina Experiment
Station, Technical Bulletin 193:110 pages.

Tsiantos, J. 1987. Transmission of bacterium Corynebacterium michiganense pv.
michiganense by seeds. J. Phytopathology 119:142-146.

Versalovic, J ., Schneider, M., De Bruijn, F. J ., and Lupski, J. R. 1994. Genomic
fingerprinting of bacteria using repetitive sequence-based polymerase chain

reaction. Meth. M01. and Cell. Biol. 5:25-40.

51

SECTION II

EFFECT OF BACTERICIDES, RESISTANT CULTIVARS, AN SAR-INDUCING
COMPOUND AND AVIRULENT STRAINS ON POPULATION SIZE AND SPREAD
OF CLA VIBACTER MICHIGANENSIS SUBSP. MICHIGANENSIS ON SEEDLING

TOMATOES IN THE GREENHOUSE

52

ABSTRACT

Treatments reduced population sizes of Clavibacter michiganensis subsp.
michiganensis among tomato seedlings in the greenhouse compared to the untreated
inoculated control in 1996 and 1998. In 1996, copper hydroxide alone or in combination
with mancozeb or streptomycin reduced pathogen populations relative to acibenzolar-S-
methyl, acibenzolar-S-methyl/copper hydroxide and avirulent C. michiganensis subsp.
michiganensis strains. Copper hydroxide/streptomycin reduced populations in
comparison to copper hydroxide/mancozeb. In 1998, treatments did not differ
Significantly in affecting population sizes. In 1997 only, treatments differed from each
other in affecting pathogen spread. Copper hydroxide mixed with mancozeb limited
spread compared to copper hydroxide mixed with streptomycin. Streptomycin sprays
were less effective than using resistant cultivars to limit pathogen spread. In the field,
inoculated control plants produced yields that were 61% (1996), 93% (1997) and 98%
(1998) of those produced by the uninoculated controls. With the exception of the
inoculated control in 1996, which resulted in yield loss, pathogen populations on
transplants used to establish field plots did not reach threshold levels of 107 CFU/g fresh
weight needed for severe symptom development or yield loss. Fruit spotting occurred

regardless of treatment and was highly variable.

53

INTRODUCTION

In 1999, the Michigan tomato (Lycopersicon esculentum Mill.) industry generated
over 27 million dollars (1). Bacterial diseases, including bacterial canker of tomato
(causal agent: Clavibacter michiganensis subsp. michiganensis), are a yearly concern for
growers in the Midwest. Infection by C. michiganensis subsp. michiganensis can result
in plant wilting, stunting, reduced yields, and plant death. Less severe disease symptoms
include marginal leaf necrosis that may be bordered by chlorosis, and fruit lesions
appearing as superficial white spots (3-6 m) that develop a necrotic center.

In the Midwest, tomato transplants are grown in local greenhouses for use in
production fields. Greenhouse transplants can appear to be healthy, yet harbor high
bacterial populations that may lead to severe disease symptoms in the field. Applications
of copper bactericides to transplants in the greenhouse can limit C. michiganensis subsp.
michiganensis population size and spread and subsequently decrease yield loss (11). The
pathogen population must be suppressed below 107 CFU/ g at the time of transplanting to
prevent yield loss in the field (11). Therefore, bactericide applications should begin in
the greenhouse rather than in the field. There are few strategies to manage bacterial
canker after symptoms appear, but growers can apply copper bactericides in the field.

Long term and continuous use of copper bactericides may lead to copper resistant
bacteria (5). While resistance to copper has not been noted to date for Clavibacter spp.
(5), we desired to develop a disease management program that is not wholly reliant on
copper bactericides. Novel products for management of bacterial diseases include those

that activate systemic acquired resistance (SAR) in plants (10) such as acibenzolar-S-

54

methyl (Actigard, Novartis Crop Protection, Inc., Greensboro, NC) (13). Tomato
cultivars moderately resistant to bacterial canker have been developed that exhibit
reduced foliar blight and yield losses compared to a susceptible cultivar (l9). Attempts
have been made to use close relatives of plant pathogens as biological control agents. In
a study by Frey et al. (7), avirulent mutants of Pseudomonas solanacerum (causal agent of
wilt in tomato) were able to establish themselves and persist in host tissue, and when
challenged with virulent strains, reduced disease symptoms. The objective of this study
was to compare the effectiveness of different managementschemes including the use of
resistant cultivars, an SAR-inducing compound, and avirulent C. michiganensis subsp.
michiganensis strains along with standard bactericides in reducing pathogen populations
and spread among greenhouse tomato seedlings. C. michiganensis subsp.. michiganensis
populations in the field and resulting yields were also of interest.
MATERIALS AND METHODS

Culture of virulent and avirulent strains. The virulent C. michiganensis subsp.
michiganensis strain used was a rifampicin-resistant mutant of a rep-polymerase chain
reaction (PCR) fingerprint type B strain isolated in 1987 from tomato fruit from a farm in
northeastern Ohio. This strain has a BOX-PCR DNA fingerprint that is different from C.
michiganensis michiganensis strains commonly encountered in Michigan (15). The
spontaneous rifampicin-resistant mutation was selected by plating cells from a nutrient
broth yeast extract (N BY) broth culture modified by omitting glucose (MNBY) (15) on
MNBY agar containing 50 pg of rifarnpicin per ml. The strain was tested for

pathogenicity to tomato by clipping the petioles of 10- to 14-day-old seedlings (cultivar

55

H8704; H. J. Heinz Co., Pittsburgh, PA) with scissors dipped in a suspension of the
strain, which had been grown in MNBY broth, centrifuging at 1,564 x g for 10 min at
room temperature in a GSA rotor in a Sorvall RC5C centrifuge (DuPont Co.,
Wilmington, DE), resuspending in sterile distilled water, and adjusting
spectrophotometrically to 3 x 108 CF U/ml. In all years, inoculated test plants developed
severe disease symptoms including stunting, wilting, and death.

The avirulent C. michiganensis subsp. michiganensis strains used as avirulent
strain treatments were streptomycin-resistant mutants of BOX-PCR DNA fingerprint type
A. Strain 40S was isolated from fresh market tomato stem tissue in 1988, and strain 3028
was isolated from a processing tomato fruit lesion in 1994; both samples were obtained in
southeastern Michigan. The spontaneous streptomycin-resistant mutation was selected by
plating cells from MN BY broth culture on MNBY containing 50 ug of streptomycin per
ml.

The virulent C. michiganensis subsp. michiganensis strain used in the greenhouse
experiments was prepared by inoculating two 5-ml broth culture tubes of MNBY
containing 100 ug of rifampicin per ml and incubating them for 48 h at room temperature
with shaking at 190 rpm. These cultures were used to inoculate 1 liter of MNBY that was
incubated for 48 h at 25°C with shaking at 75 rpm. The culture was centrifuged and the
pellets were resuspended and combined in sterile distilled water to a final volume of 1
liter; the final concentration was 5 x 108 CFU/rnl. The avirulent strains were prepared as
described above, except MNBY broth contained 100 ug of streptomycin per ml. The

suspensions were kept on ice and applied within three hours using a C02 backpack

56

sprayer with two flat-fan 8002 nozzles (Teejet, Chicago) (1996, 1997) or a single nozzle
(1998) which were operated at 2.8 kg/cm2 delivering approximately 748 liter/ha.

Experimental design and treatments. On 13 and 14 May 1996, 24 March 1997,
and 23 March 1998, plastic plug sheets (52.5 cm x 26.5 cm x 4.0 cm) each containing 288
cells filled with soilless medium were individually seeded with the tomato cultivars
H8704, H9144, and H70214 (H. J. Heinz Co., Pittsburgh, PA; 1996, 1997) and cultivar
Mountain Spring (Novartis Seeds, Inc., Gilroy, CA; 1998) and germinated for 3 days in a
walk-in germination chamber in a commercial greenhouse in southwestern Michigan.
Data collected in 1996 and 1997 are included in this study to strengthen the importance of
this research, although were not a part of the work performed to complete the masters
degree. Plug sheets were then transported to a commercial polyethylene greenhouse
(approximately 12.2 m x 29.3 m) in Stockbridge, Michigan, where they were placed on
overturned plastic flats on the earthen greenhouse floor, that was covered with a black
woven polyethylene groundcover. Flats were arranged in blocks, each consisting of 12
plug sheets (1996, 1997) or 15 plug sheets (1998) (Fig. 1). Seedlings were watered
overhead as required, and irrigated with 100- or ZOO-ppm N and K20 fertilizer solution
(Peter’s 20-20-20; Grace-Sierra Horticultural Products Co., Milpitas, CA) as needed.
Greenhouse temperatures were monitored and average maximum and minimum
temperatures of 259°C (day) and 116°C (night), respectively were recorded. The
average relative humidity was 95.4%.

Each year the plug sheets were arranged in a block design, with four replicates per

57

 

Figure l. Diagrams of treatment blocks with shaded areas serving as a buffer between
blocks and each rectangle representing a 52.5 x 26.5 cm tray containing 288 tomato
seedlings. In 1996-1997 (A), seedlings were sampled from each block following the
diagonal (regions a to d) from the Clavibacter michiganensis subsp. michiganensis
inoculum focus (*). In 1998 (B), seedlings were sampled from each block following the
diagonal (regions a to e) between the C. michiganensis subsp. michiganensis inoculum
foci (*).

58

treatment. The inner four (2 X2) plug sheets (1996, 1997), and the inner six (2X3) plug
sheets (1998) of each block were targeted for treatment, with the surrounding flats
considered as buffers between blocks (Fig. l). Fungicides were applied as needed to
manage nontarget pathogens including Bottytis cinerea (1997, 1998), A lternaria solam'
(1998), Sclerotinia sclerotiorum (1998) and Oidiopsis sicula (1998).

The following treatments were investigated: (i) copper hydroxide (Kocide 40DF,
3.0 g a.i./liter; Griffin LLC, Valdosta, GA); (ii) copper hydroxide (Kocide 40DF, 3.0 g
a.i.lliter) mixed with mancozeb (Dithane F-45, 2.3 ml a.i.lliter; Rohm and Haas Co.,
Philadelphia); (iii) copper hydroxide (Kocide 40DF, 3.0 g a.i.lliter) mixed with
streptomycin (Agri-mycin 17, 0.25 g a.i./liter; Novartis Crop Protection, Inc. Ag.
Products, Greensboro, NC) (1996 and 1997 only); (iv) SAR-inducing compound
acibenzolar-S-methyl (Actigard 50WDG, 0.1 g a.i./liter; Novartis Crop Protection, Inc.
Ag. Products, Greensboro, NC); (v) acibenzolar-S-methyl (Actigard 50WDG, 0.1 g
a.i./liter) mixed with copper hydroxide (Kocide 40DF, 3.0 g a.i.lliter) (1996 only); (vi)
resistant cultivar H70214 (H. J. Heinz Co.; 1996 and 1997 only); (vii) resistant cultivar
H9144 (H. J. Heinz Co.; 1996 and 1997 only); (viii) avirulent streptomycin-resistant C.
michiganensis subsp. michiganensis strain 3028 (1996 and 1997 only); (ix) avirulent
streptomycin-resistant C. michiganensis subsp. michiganensis strain 408 (1996 and 1997
only); (x) streptomycin (Agri-mycin 17, 0.25 g a.i.lliter) (1997 and 1998 only); and (xi)
no treatment (control). Untreated, uninoculated plants of each cultivar were grown in a
separate greenhouse under environmental and cultural conditions similar to those of the

inoculated treatments. In 1998, three sets of four replicates of untreated, inoculated

59

controls were included to provide sufficient plants for subsequent field treatments.

Treatment sprays were applied using a C02 backpack sprayer with two flat-fan
8002 nozzles (Teejet, Chicago) (1996, 1997) or a single nozzle (1998) which were
operated at 2.8 kg/cm2 delivering approximately 748 liter/ha. Sprays were initiated when
the first true leaves of the seedlings were visible and 1 to 2 days prior to inoculation with
the virulent rifampicin-resistant C. michiganensis subsp. michiganensis. Subsequent
sprays were applied every 5 days (10 June to 5 July 1996; 22 April to 17 May 1997; 25
April to 4 June 1998) until the seedlings were removed from the greenhouse and planted
in the field.

Inoculation. On 11 June 1996, 23 April 1997, and 27 April 1998, seedlings
(cultivar H8704) were inoculated by misting the leaves with a virulent rifampicin-
resistant C. michiganensis subsp. michiganensis suspension (prepared as described
above) using a Preval pressurized sprayer (Precision Valve Corporation, Yonkers, NY).
The first true leaf of each individual seedling was then removed by clipping the petiole
close to the stem with scissors dipped in the same C. michiganensis subsp. michiganensis
suspension. Inoculated seedlings were incubated overnight on a laboratory bench in
loosely closed plastic bags to maintain high humidity. To initiate infection in the
commercial greenhouse, 16 seedlings with their soilless plugs and roots intact were
removed from the comer closest to the ‘a’ site (1996, 1997), or from each of the two
opposite corners of each treatment block (1998) and replaced with the same number of
inoculated seedlings to establish an inoculum focus(i) (Fig. 1).

Sampling of greenhouse seedlings. At 28 days (1996), 29 days (1997), and 42

60

days (1998) after inoculation, five to eight seedlings were removed from four regions (a
to d; 1996, 1997) or five regions (a to e; 1998) within each block along the diagonal from
the inoculum focus(i) (Fig. l). The inoculated plants used to initiate disease within each
treatment block were not included in the foliar samples; they were dead by the time of
sampling. Samples from each region were stored individually in plastic bags in a cooler
at 4°C for a maximum of 3 days; each sample was individually processed. Shoots and
leaves were chopped with a sterile straight razor and homogenized for 2 min in a Lab-
Blender 400 stomacher (Tekmar Co., Cincinnati, OH) using sterile phosphate buffer (0.05
M, ph 7.4; 2 ml/g of plant tissue) amended with Tween-20 detergent (0.02%). One mL of
the plant extract was then stored in 0.5 ml of 40% glycerol and frozen at -20°C until
further processing.

Each sample was thawed on ice and population sizes of the virulent pathogen
were estimated from lO-fold serial dilutions spread on MNBY agar containing 100 ug of
rifampicin per ml. Colonies were counted after 4 to 7 days. The lower limit of detection
was 30 CFU/g of tissue. To verify that these isolates had the same BOX-PCR DNA
fingerprint type as the inoculum, 18 (1996), 10 (1997), and 23 colonies (1998), were
chosen for BOX-PCR DNA fingerprinting. In 1996, 15 colonies were the rifampicin-
resistant B type, while the others were .not C. michiganensis subsp. michiganensis. In
1997, none of the chosen colonies were C. michiganensis subsp. michiganensis. In 1998,
all colonies were the rifampicin-resistant B type.

Untreated, uninoculated plants were sampled by haphazardly removing four

seedlings from one flat (1996, 1997) or six seedlings from each cultivar (1998) and

61

processing as previously described. Serially diluted homogenates were plated on MN BY
agar containing 100 pl of rifampicin per ml. There were no colonies resembling C.
michiganensis subsp. michiganensis in any year.

Statistical analysis of greenhouse populations. P0pulation data were analyzed
by ANOVA as a completely randomized split-plot design with the main plots as the
block, which differed by treatment, and position as sub-plots. The 1996 and 1997 data
were transformed using Y=ln(CFU+1) to achieve normality, as determined by analyzing
residuals using the SAS procedure, Proc Univariate (SAS Institute, Cary, NC). The
assumption of equality of variances was examined using Levene’s Test (22). Because the
1998 variances were not equal, an AN OVA of ranked population sizes was conducted (4)
(Proc Rank and Proc GLM [SAS Institute, Cary, NC]). This analysis is similar to other
non-parametric procedures for less complex designs (4). Although the analyses is
affected by unequal variances, this effect should be minimal, especially compared with a
parametric analysis. The following linear model was used for all ANOVA procedures:
mean + treatment + position + treatment x position + block nested within treatment +
error; block nested within treatment was used as the error term for treatment.

For all years, linear contrasts were used to examine the differences between
treatments and the effects of treatment on the relationship of populations with position in
the experimental blocks. The following contrasts were used to examine the differences in
treatments for 1996: (i) inoculated susceptible control versus all other treatments; (ii)
inoculated resistant cultivars versus all other treatments; (iii) copper hydroxide alone or

mixed with streptomycin or mancozeb versus acibenzolar-S-methyl alone or with copper

62

hydroxide, and avirulent strain treatments; (iv) acibenzolar-S-methyl alone or with copper
hydroxide versus avirulent strain treatments; (v) copper hydroxide alone versus copper
hydroxide mixed with streptomycin or mancozeb; (vi) copper hydroxide/streptomycin
versus copper hydroxide/mancozeb; (vii) acibenzolar-S-methyl versus acibenzolar-S-
methyl/copper hydroxide (1996 only); (viii) avirulent strains 403 versus 3028. The
effects of treatment on the relationship of populations with position in the experimental
blocks (interaction of treatment and position) was examined using contrasts similar to
those above for 1997, except the streptomycin treatment was considered a positive
control.

The following contrasts were used to examine the differences in treatments for
1998: (i) copper hydroxide alone or mixed with mancozeb versus the inoculated control;
(ii) copper hydroxide alone versus copper hydroxide/mancozeb; (iii) acibenzolar-S-
methyl versus inoculated control; (iv) streptomycin versus inoculated control; (v) copper
hydroxide versus acibenzolar-S-methyl and streptomycin: (vi) acibenzolar-S-methyl
versus streptomycin. Because these contrasts were not independent, and not entirely a-
priori comparisons, Bonferonni’s correction was used to adjust a to 0.008 (0.05/6); only
contrasts with a p-value less than 0.008 were considered significant.

Field Study. On 16 and 17 July 1996, 28 May 1997, and 10 June 1998, 36
seedlings were randomly selected from the center (Fig. 1, regions b and c; 1996, 1997) or
(Fig. 1, region c; 1998) of each treatment block and planted in a randomized complete
block design with 4 replications at the Botany and Plant Pathology Research Farm, East

Lansing, Mich, in sandy loam (1996, 1998) or clay loam (1997). The fields had been

63

previously planted to sweet com (1996), or rye (1997, 1998). Each treatment plot
consisted of 12 seedlings planted into each of three 3.6-m-long rows spaced 1.5 m apart
with 30.5 cm between plants within rows. The experimental sites were fertilized, and
weeds were managed according to standard commercial practices.

In 1996 and 1997, to prevent spread of the pathogen from nearby inoculated
plants, uninoculated control plants received streptomycin (Agri-mycin 17, 0.187 kg
a.i.lha) applications weekly using a pneumatic handsprayer with a single nozzle operated
at 2.8 kg/cm2 delivering 748 liters/ha. In 1998, streptomycin was applied to a plot of
uninoculated plants just outside the experimental field using a C02 backpack sprayer with
a 8003LP nozzle operated at 2.8 kg/cmz, delivering 467 liters/ha. In this year the
uninoculated plants randomized within the experimental field did not receive
streptomycin applications. Each year, all treatments were sprayed, as needed, using the
same method as described above, with the fungicide chlorothalonil (Bravo Weather Stik,
2.5 kg a.i./ha; Zeneca Ag Products, Wilmington, DE) to prevent disease caused by fungal
pathogens.

Field sampling of fruit and foliage. On 2 October 1996, 8 September 1997, and
23 September 1998, plant stand counts were recorded for the middle 3.6-m-long row, and
total yield, and incidence of fruit spotting were recorded in a single harvest for the five
innermost plants of the middle row. Foliage from plants was sampled on 26 September
1996, 4 September 1997, and 21 September 1998 at random from five to six central
plants in the center row of each treatment plot and processed as previously described to

determine the pathogen population sizes. To verify that the field isolates were identical to

64

the one used to inoculate the seedlings in the greenhouse, 44 (1996), and 106 (1998)
colonies resembling C. michiganensis subsp. michigancnsis were chosen for BOX-PCR
DNA fingerprinting. In 1996, 21 colonies were rifampicin-resistant B types, 3 were
rifampicin-resistant C types, and 20 were not the pathogen. In 1998, 103 colonies were
the rifampicin-resistant B type, 1 was a rifampicin-resistant C type, and 2 were not C.
michiganensis subsp. michiganensis.

In 1998, diseased fruit were surface sterilized with 70% ethanol (v/v), and a
sterile toothpick inserted just below the surface of bird’s eye lesions and streaked onto
MNBY containing 100 pl of rifampicin per ml. F ourty-seven colonies resembling C.
michiganensis subsp. michiganensis were subjected to BOX-PCR DNA fingerprinting
and determined to be the rifampicin-resistant B type, except for one colony that was a
rifampicin-resistant C type.

Statistical analysis of field measurements. The 1996 and 1997 field data were
analyzed as a randomized complete block design. Total fruit weight was transformed
using Y = (X+1)”. The transformed data were normally distributed with equal variances.
Stand count data were not normally distributed, the variances were not equal, and the data
could not be transformed to meet these assumptions. Therefore, ranks of this variable
were used in all subsequent analyses. The following contrasts were used to examine the
differences in treatments for 1996: (i) susceptible and resistant inoculated versus
susceptible and resistant uninoculated controls; (ii) susceptible control, inoculated or not
versus all other treatments; (iii) resistant cultivars, inoculated or not versus all other

treatments; (iv) inoculated susceptible control versus all other chemical, acibenzolar-S-

65

methyl alone or with copper hydroxide and avirulent strain treatments; (v) uninoculated
susceptible control versus all other chemical, acibenzolar-S-methyl alone or with copper
hydroxide and avirulent strain treatments; (vi) inoculated resistant cultivars versus all
chemical, acibenzolar-S-methyl alone or with copper hydroxide and avirulent strain
treatments; (vii) uninoculated resistant cultivars versus all chemical, acibenzolar-S-
methyl alone or with copper hydroxide and avirulent strain treatments; (viii) the
interaction between inoculation and control (positive vs. negative); (ix) copper hydroxide
alone or mixed with streptomycin or mancozeb versus acibenzolar-S-methyl alone or with
copper hydroxide and avirulent strain treatments; (x) acibenzolar-S-methyl alone or with
copper hydroxide versus avirulent strain treatments; (xi) copper hydroxide versus copper
hydroxide mixed with streptomycin or mancozeb; (xii) copper hydroxide/streptomycin
versus copper hydroxide/mancozeb; (xiii) acibenzolar-S-methyl vs. acibenzolar-S-
methyl/copper hydroxide (1996 only); (xiv) avirulent strain 408 versus 3028. In 1997,
similar contrasts were used, except the streptomycin treatment was considered a positive
control, and used as such for the following contrasts: (i) susceptible and resistant
inoculated versus susceptible and resistant uninoculated controls; (ii) resistant cultivars,
inoculated or not versus all other treatments.

Plant stand and yield data for the 1998 field plot were not statistically analyzed
due to additional treatments applied in the field that were not included in the 1996 and

1997 studies and therefore not reported in this paper.

66

RESULTS

Impact of bactericides, resistant cultivars, an SAR-inducing compound, and
avirulent strains on pathogen populations and spread among transplants in the
greenhouse. Symptoms observed on seedlings that were directly inoculated with bacteria
and introduced into the treatment blocks included cankers, firing, wilt, and death. In all
years, C. michiganensis subsp. michiganensis populations were detected in all inoculated
treatment blocks but not in the uninoculated controls.

In 1996, treatments significantly (P < 0.001) reduced pathogen populations
compared to the inoculated susceptible control (Table 1). However, pathogen population
sizes on treated plants were the same as on the inoculated resistant culitvars (P = 0.409).
Copper hydroxide alone or mixed with mancozeb or streptomycin significantly reduced
pathogen populations compared to acibenzolar-S-methyl alone or with copper hydroxide
and avirulent strain treatments (P = 0.054). The addition of streptomycin or mancozeb to
copper hydroxide did not enhance control (P = 0.377), yet copper hydroxide/streptomycin
significantly reduced population sizes compared to copper hydroxide/mancozeb (P =
0.035). There was no apparent advantage in adding copper hydroxide to acibenzolar-S-
methyl since pathogen populations were comparable (P = 0.084). Acibenzolar—S-methyl
alone or mixed with copper hydroxide reduced pathogen population sizes compared to the
avirulent strain treatments (P = 0.019). The avirulent strain treatment 40S did not differ
from 3028 in limiting pathogen populations (P = 0.953). In 1997, treatments did not

affect pathogen population sizes (P = 0.333).

67

Table 1. Summary of contrast results comparing treatments applied to tomato seedlings
in the greenhouse when inoculated with CIavibacter michiganensis subsp. michiganensis

in relation to pathogen population size in 1996.

 

 

1996
Contrast F P
Inoculated susceptible control vs. all other treatments3 19.77 <0.001"‘b
Inoculated resistant cultivars vs. all chemical, SAR-inducing
compound“ and avirulent strain treatments 0.70 0.409
Copper hydroxide alone or mixed with streptomycin or
mancozeb vs. SAR-inducing compound and avirulent strain
treatments 4.03 0.054*
SAR-inducing compound vs. avirulent strain treatments 6.12 0019*
Copper hydroxide alone vs. copper hydroxide mixed with
streptomycin or mancozeb 0.80 0.377
Copper hydroxide/streptomycin vs. copper hydroxide/mancozeb 4.88 0035*
Acibenzolar-S-methyl vs. acibenzolar-S-methyl/copper
hydroxide 3. 1 9 0.084
Avirulent strain treatments (408 vs. 3028) 0.00 0.953

 

“ Copper hydroxide (Kocide 40DF, 3.0 g a.i./liter); mancozeb (Dithane F-45, 2.3 ml
a.i.lliter); streptomycin (Agri-mycin 17, 0.25 g a.i./liter); acibenzolar-S-methyl (Actigard

50WDG, 0.1 a.i./liter); resistant cultivars H70214 and H9144. (H. J. Heinz Co.,

Pittsburgh, PA); avirulent streptomycin-resistant C. michiganensis subsp. michiganensis
strains 40S and 302S. Number of applications (if pertinent) were 1x/5 days = five sprays.

° * Indicates the contrast is significant at P s 0.05.
“ Acibenzolar—S-methyl and acibenzolar-S-methyl/copper hydroxide.

68

In 1998, copper hydroxide alone or mixed with mancozeb significantly reduced
populations relative to the inoculated control (P < 0.001) as did acibenzolar-S-methyl (P
< 0.001) and streptomycin (P < 0.001) treatments (Table 2). However, adding mancozeb
to copper hydroxide did not offer a significant benefit compared to copper hydroxide
alone (P = 0.265). Copper hydroxide was comparable to acibenzolar-S-methyl and
streptomycin (P = 0.109) and acibenzolar-S-methyl was comparable to streptomycin (P =
0.509) in limiting population sizes.

Treatments significantly affected pathogen spread (P = 0.001) in 1997 only.
Copper hydroxide mixed with mancozeb limited spread compared to copper hydroxide
mixed with streptomycin (P < 0.001), yet there was no difference in spread between
copper hydroxide alone versus copper hydroxide mixed with mancozeb or streptomycin
(P = 0.074). Streptomycin was less effective than the inoculated resistant cultivars in
limiting pathogen spread (P = 0.004).

Impact of bactericides, resistant cultivars, an SAR-inducing compound, and
avirulent strains on field productivity. At the time of transplanting, the maximum
bacterial populations in the untreated susceptible controls used to establish field plots at
regions b and c (1996, 1997; Fig. 1), or at region c (1998; Fig. 1) were 4.1 X 107 (1996),
6.1 X 10’ (1997), and 1.4 X 105 (1998) (Fig. 2). All treatments limited pathogen
populations at regions b and c to <8.2 X 10“ CFU/g (1996) and <42 x 105 CFU/g (1997)
of tissue. In 1998, all treatments had undetectable pathogen populations at region c (Fig.
3).

In all years, plants showed foliar symptoms of bacterial canker at the end of the

69

Table 2. Summary of contrast results comparing treatments applied to tomato seedlings
in the greenhouse when inoculated with Clavibacter michiganensis subsp. michiganensis

in relation to pathogen size in 1998.

 

 

Contrast F P
Copper hydroxidea alone or mixed with mancozeb“ vs.

inoculated control 74.95 <0.001"‘c
Copper hydroxide vs. copper hydroxide/mancozeb 1.31 0.265
Acibenzolar-S-methyld vs. inoculated control 30.88 <0.001*
Streptomycin“ vs. inoculated control 37.07 <0.001*
Copper hydroxide vs. acibenzolar-S-methyl and streptomycin 2.81 0.109
Acibenzolar-S-methyl vs. streptomycin 0.45 0.509

 

° Copper hydroxide (Kocide 40DF, 3.0 g a.i./liter). Number of applications for all

treatments were 1x/5 days = nine sprays.

b Copper hydroxide (Kocide 40DF, 3.0 g a.i.lliter) mixed with mancozeb (Dithane F-45,

2.3 ml a.i.lliter).

“ * Indicates the contrast is significant at P < 0.05.

“ Acibenzolar-S-methyl (Actigard 50WDG, 0.1 g a.i./liter).
“ Streptomycin (Agri-mycin 17, 0.25 g a.i.lliter).

7O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

hoctlatod susceptble control C l I aid.
910 . 01° 7
a 8 a 8
O ' 0 5
3 ‘ a 4
- o r o
C b c d
Sampleslh
WWW
.10 . “1o
5 a E a
'6 6 o s
3 ‘ a 4
2 2
3 o s o
I b c d
Sampleslto
Mow-am
10 10
£8 ‘ 2 a .
3 ‘ ‘ g 5.
2“ 24-
32 - 3 2 .
o- o_
I b g d
Samples":
403
10 1o ,
a
E “ g a
o ‘ e
2 4 a ‘
8’ 2 2 E
- o s , I
a h e d . b c d I
".1“ . Acibenzolaté-methyerpper hydroxrde ‘
10 j .
9 8 1 10 ’
E e 8 !
0 ‘ a 3, i
3 4 q 4‘ .
2 ' 2* El 1
5 o E] 5 o D I: i
a b c d a b c d I
“mph“ Samplosm l
Streptomycin
10 l
i ‘ ~
6
e 4 l
2
I b c d '
ammun- {

 

Figure 2. Clavibacter michiganensis subsp. michiganensis populations (CFU/ g) on
treated or untreated tomato seedlings in the greenhouse prior to transplanting in the
field in 1996 (c1) and 1997 (I) when sampled from sites adjacent to C. michiganensis
subsp. michiganensis inoculum focus (a), the center of the treatment block (b,c), and
the point farthest from the inoculum focus (d). Error bars represent standard error.

71

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 Inoculated control 1 10 7 Inoculated control 2
2’ 3 2 8
a
g 6 1:5 6
2 4 2 4
3 2 5 2
o O
. b SampIe slte d . . b SampIe site d .
Inoculated control 3 Copper hydroxide
1o , 10 .
a a
o '5 0 °
3 4 3 4
3 2 5’ 2
o o
a b c d e a b c d e
Sample site Sample the
Copper hydroxidelmancozeb Aclbenzolar-S-methyl
1o 10 *
'5’ a 2 ll
3 6 E t
a 4 a 4
E 2 5' 2
0 o
3 5 ¢ 9 ‘ a b c d e
Sample alte Sample site
10 Streptomycin
8
B
: 4
g 2
o
a b c d e
Sample site

 

 

 

Figure 3. C lavibacter michiganensis subsp. michiganensis populations (CFU/ g) on
treated or untreated tomato seedlings in the greenhouse prior to transplanting in the field
in 1998 when sampled from sites adjacent to C. michiganensis subsp. michiganensis
inoculum foci (a, e), the center of the block (c), and the points in between (b, (1). Three
untreated controls were included in the study to supply enough untreated plants for
subsequent field treatments. Error bars represent standard error.

72

growing season including necrosis of the outer leaf edges, or “firing”, and brown to tan
lesions on peduncles. Final foliar populations ranged from 9.3 X 107 to 1.8 X 109 (1996),
1.4 X 107 to 3.9 X 108 (1997), and 1.6 X 10“ to 3.1 X 108 (1998) CFU/g oftissue. The
uninoculated untreated plants, with applications of streptomycin in 1996 and 1997 only.
had populations of7.1 X 105 to 9.1 X 10“ (1996), 1.7 X 107 to 3.6 X 107 (1997), and 4.2 X
104 to 5.0 X 106 CFU/g oftissue (1998).

In 1996, fewer plants survived in the inoculated untreated susceptible (9.5 out of a
maximum of 12) compared to the other treatments (11.0) (P = 0.005). In 1997, plant
stand was reduced in all treatments due to hot, dry conditions during establishment of
transplants in the field. Plant survival in the inoculated susceptible control (8 out of a
maxiumum of 12) did not differ significantly from the other treatments (7.3 to 10.3
plants) (P = 0.411). In 1998, all plants survived in the inoculated susceptible control.
Plant survival ranged from 8.5 to 12 in the other treatments.

Inoculated control plants produced yields that were 61% (1996), 93% (1997)
(Table 3), and 98% (1998) (data not shown) of those produced by the uninoculated
controls. In 1996, the yield from the inoculated susceptible control was significantly
reduced compared to the treatments included in this study (P = 0.021). In 1997, the yield
from the inoculated susceptible control did not differ significantly from the other
treatments (P = 0.457). However, the yield from plants treated with copper hydroxide
alone was significantly reduced compared to copper hydroxide mixed with streptomycin
or mancozeb (P = 0.005). The yields of the resistant cultivars (inoculated and

uninoculated) were significantly less than the other treatments in 1996 (P = 0.019) but not

73

”2 Wm .5833 33:3 8938 vows—32882

 

 

0.3 we SAENEIV .8323 898mm: vows—30:83
Q: m .w @395 33:3 28388 @2232:
”2 34 A3823 33:3 8min“: Ban—:85
v.2 5w wow 88?. ace—83¢.
_.0_ ad mmom £88 828m><
ed 0286»: SmacfiifioEm-cm_oN:o£o<

Wm _ 9w _388-m-§omco£o<
odm a... . 80%8838m
2: ed 83883880286»: 8&qu
92 _.m 238583386»: 5880
0.. L m6 028.6%: Sacco
v.2 w.m 7588 288083 Baa—=8:—
o.: Wm $888 «58083 Bum—soocmcD
53 :3 «80839.?

35.... mass 23>

 

.32 can 003 8 8:98on 05 8 858:8: beam 9 380.3% 88 mmmamaewESE .mmnzm SESSMESE $83339 5:3
Ban—:88 «o: co 36388 $8803. $838 2988 :0 358838 @8388 3:58 3888 mo $883 can 22% 28m .m 03:.

74

 

 

 

S _ .o E: mwwd .86 8888 5:03.88 :8 88888

w80::8um<m .m> 308088 8 888888 83 888 8 0:08 0:58.88 8800

m Ed 8.. S _ .o 9: A- .m> +v 35:8 :8 858388 803503 8508058

88 :8 8 _ .o cod . 88805 :88 80:88

:8 88888 w80::8-m<m 88808 :8 .m> m8>530 58580: 8830882

*wmod we: * _No.o oww 350.5805 88.8

80388 :8 88888 w80::8-m<m 8880.3 :8 .m> m8>530 58880: 88388

88 cod 88 98 80880: :88 80.88

:8 .:::o:88 w80::8-m<m 808030 :8 .m> 35:8 035888 8830883

mmvd nmd * God mmfi 8808805 888 80388

:8 88888 w80::85m<m 808030 :8 .m> 35:8 035888 88388

8 _ .o a: 253: .8: 80:58: :88 80:88

:8 {8:888 w80::8-m<m 88808 :8 .m> 930 boo: 8058888 :85 :0805

8>530 035888 88388 :8 m8>530 58580: 88388:: :8 88388

350 Ed Smd N. _ ._ 8808805

8:8 :8 .m> 35:8 035888 88388:: :8 3580 035888 88388

no _ .o m8.— omod Ed m8>530 035888 :8 58:80: 88388:: .m> m8>530

930 803 888988 .555 8805 035888 :8 035888 “58880: 88388

k k k k 88.580
32 :8—

853 80: 23>

 

8:888 .m 038%

75

5:0 com: E _»fioE-m-§oN:o£om 53» 3me mm? 2586»: hanou g
.modvm an 685%? mm 58:80 32365 _.. u

.oExoGE 8308.388-m-3_owco£om Ea Rfiofiiém—ogonmoax u
.80» 8:898 35 E nous—2: Ho: «5838,?

630:5on 8888 a E 850% 2333a; p

.33 can 82 5 $9an 96 u what m5: 803 95523

.3 mcozmozmam mo 52:52 .mov can mm?“ 385 wfizmzawEuE dmnsm flatmacwifls U ESm_m2-=_o>anobm “co—33a A:

.ng35 :8 NEE .. .5 35: Ba ESE $323 6%me Eo:_\._.m S .838 2&3; Esfiméséoes 52:2“
w 85 .: screams 505335 625.: 2. mm .22 82:9 38058 625.: w 3 .38 028v: 02362 5&8.

 

 

 

 

vmvd nmd mugs—so 3932 3532: .m> Eobcofiobm
93d 85 23 SS 38m 2 m9; 25% 62.52
mood o _ .o .oExEPE Homaoo\_>508-m¢m_o§onmom .m> _mfioE-m¢m_0N:o£o<

086 mm; wood cod no~8§~5o2§6§ 6&8 .m> onEoEomeExEuE .5995
$36 mod chad cod @3858 5 538935 53> 3me QExBPE 5&8 .m> 02x06»; 5&qu
mnod mva wmwd mod 3:253: E86 Eo_3_>m .m> 952.88 wcmozvfiilam
& k E K Bah—BU

baa—
Efia 35 22>

 

3:568 .m 29$

6
7

(Table 4). In 1996, there were no diseased fruits in the uninoculated resistant cultivars or
in 1997 (P = 0.184). However, the yield of the inoculated resistant control was
significantly less (P = 0.021) in 1996 and significantly greater (P = 0.038) in 1997 than
all other treatments. In 1998, the yield from the uninoculated control was 26.9 kg/S
plants and the yield from the treatments ranged from 26.4 to 32.7 kg/S plants.

The number of fruit with spots was highly variable among years with incidence
much lower in 1996 (2.8%) (Table 4) and 1998 (3.3%) (data not shown) than in 1997
(47.9%) the susceptible control compared to s 1.6% when these controls were inoculated.
In most treatments, <1% of the fruit were diseased. However, in 1997, the overall
incidence of fruit with spots was high (47.9%). Incidence of fruit spots was 30.8% versus
s 16.2% in the inoculated susceptible control and resistant cultivars, respectively. In the
uninoculated resistant cultivars and the susceptible control, the incidence of fruit with
spots was s 8.9%. While treatment with copper hydroxide resulted in fruit spotting of
47.9%, all other treatments resulted in fruit spotting incidence of 11.7% to 22.4%. In
1998, no fruit were diseased in the uninoculated untreated plot compared to 2.4% in the
inoculated untreated plot (data not shown). The incidence of fruit with spots was s 0.3%
in all treatments with the exception of copper hydroxide/mancozeb (3.3%).

DISCUSSION

A' C. michiganensis subsp. michiganensis population of 107 CF U/ g of tissue or
higher occurred on the inoculated untreated susceptible seedlings at the end of the
greenhouse growing cycle (just prior to planting in the field) in 1996 and was associated

with development of severe disease symptoms in the field such as plant stunting and yield

77

Table 4. Incidence of fruit spotting on tomato plants in the field; seedlings inoculated or
not inoculated with CIavibacter michiganensis subsp. michiganensis and subjected to
spray treatments in the greenhouse in 1996 and 1997.

 

 

 

Diseased fruit (°/o)b

Treatment“ 1996 1997
Uninoculated controlc 0.0 8.9
Inoculated control 1.6 30.8
Copper hydroxide 0.2 47.9
Copper hydroxide/mancozeb 2.8 22.4
Copper hydroxide/streptomycin 2.0 1 1.7
Streptomycin 'Ad 1 7. 1
Acibenzolar-S-methyl 0.8 20.2
Acibenzolar-S-methyl/copper hydroxide 0.3 ‘A

Avirulent strain 3028 0.0 17.2
Avirulent strain 408 1.8 19.1
Inoculated resistant cultivar (H70214) 0.3 16.2
Inoculated resistant cultivar (H9144) 0.6 5.3
Uninoculated resistant cultivar (H70214)° 0.0 7.9
Uninoculated resistant cultivar (H9144)c 0.0 6.5

 

“ Copper hydroxide (Kocide 40DF, 3.0 g a.i.lliter); mancozeb (Dithane F-45, 2.3 ml
a.i.lliter); streptomycin (Agri-mycin 17, 0.25 g a.i.lliter); acibenzolar-S-methyl (Actigard
50WDG, 0.1 a.i./liter); resistant cultivars H70214 and H9144 (H. J. Heinz Co.,
Pittsburgh, PA); avirulent streptomycin-resistant C. michiganensis subsp. michiganensis
strains 3028 and 408. Number of applications (if pertinent) were 1x/5 days = five sprays
in 1996 and 1997.

b By weight.

° Transplants grown in a separate greenhouse.

d Treatment was not included in that particular year.

78

loss (39%). A previous study indicated that susceptible seedlings with C. michiganensis
subsp. michiganensis populations of 107 CFU/g of tissue or higher were associated with
systemic infection symptoms in the field including plant stunting, yield loss, and death
(11). Treatments included in this study suppressed pathogen populations on susceptible
transplants below the threshold level (107 CFU/g of tissue) and yield losses were not
observed.

In 1997 and 1998, pathogen populations at the end of the greenhouse growing
cycle did not exceed 6.1 X 105 CFU/g of tissue, even in the inoculated untreated
susceptible seedlings. Consequently, yield losses in the field were not observed among
treatments or the controls, verifying our theory that transplants placed in the field with
low pathogen populations will produce commercially acceptable yields under Michigan
growing conditions. One exception occurred in 1997 with copper hydroxide that had a
very low plant stand count (3 out of 12 plants) in the field shortly after transplanting in
one of the four blocks due to unfavorable field conditions and likely contributed to the
observed decreased yield.

Treatments significantly reduced pathogen population sizes among tomato
seedlings in the greenhouse compared to the untreated control in 1996 and 1998. In 1996
only, the conventional bactericides including copper hydroxide alone or mixed with
mancozeb or streptomycin reduced pathogen populations among seedlings in the
greenhouse compared to acibenzolar-S-methyl alone or mixed with copper hydroxide.
However, acibenzolar-S-methyl alone or mixed with copper hydroxide was more

effective than the avirulent strains. Streptomycin was included in our trial as a standard

79

but cannot be recommended for use in Michigan because the label does not specify
greenhouse as an application site.

In 1996 only, yields from the inoculated and uninoculated resistant cultivars were
significantly reduced compared to the other treatments. These plants did not exhibit
significant disease symptoms and the pathogen population did not exceed the 107 CFU/ g
of tissue threshold. The resistant cultivars used were late maturing and may have been
negatively affected by the shortened growing season in 1996 (66 days) compared to that
of 1997 (99 days). In 1997, these same inoculated cultivars yielded more than the
chemical, acibenzolar-S-methyl and avirulent strain treatments.

In our study, it was not beneficial to add mancozeb to copper hydroxide to reduce
pathogen spread in the greenhouse, as suggested in a previous study (11). In other
studies, copper hydroxide/mancozeb has been shown to enhance control of bacterial
diseases and reduce epiphytic bacterial populations compared with using copper alone
(12, 17). In 1997 only, pathogen spread was limited among transplants treated with
copper hydroxide/mancozeb compared to plants treated with copper
hydroxide/streptomycin and among inoculated resistant plants compared to plants treated
with streptomycin.

Other studies have shown positive results with acibenzolar-S-methyl for
management of bacterial diseases. In a field study, acibenzolar-S-methyl provided better
control of bacterial spot (causal agent: Xanthomonas campestris pv. vesicatoria) (14), and
bacterial speck (causal agent: Pseudomonas syringae pv. tomato) (16) on tomato than

copper hydroxide mixed with either maneb or mancozeb, respectively. In a greenhouse

80

tobacco study, acibenzolar-S-methyl alone, or mixed with copper oxychloride was more
effective than copper oxychloride alone in reducing chlorosis and necrosis caused by
Pseudomonas syringae pv. tabaci tox + (3). In tobacco fields, disease (causal agent, P.
syringae pv. tabaci tox +) symptoms were reduced when seedlings were treated with
copper oxychloride only in the seed bed (9.3%), or treated with acibenzolar-S-methyl
alone (2.7%) or in combination with copper oxychloride in the seed bed and once in the
field (4.7%), compared to the untreated control (72%) (3).

Final foliar field populations at the end of the growing seasons ranged from 1.4 x
107 to 1.8 X 10°; the uninoculated controls had slightly lower pathogen populations (4.2 X
104 to 3.6 X 107). All treatments, including the uninoculated control, exhibited leaf
margin necrosis. Chang et al. determined that relatively high (107 to 109 CF U/g fresh
weight) leaf surface populations of C. michiganensis subsp. michiganensis can occur at
least 2.7 m from the focus of infection resulting in secondary spread and relatively minor
disease symptom development (2). The field populations observed in our study exceeded
the leaf surface population of C. michiganensis subsp. michiganensis of 106 to 107 CF U/g
fresh weight that Chang et al. determined is necessary before symptoms of secondary
infection, including spotted fruit and firing of leaflets, occur (2). In Michigan, bacterial
canker symptoms such as necrosis of the leaf margin occur in mid to late season but are
seldom associated with economically significant losses (20, 21).

“Bird’s-eye” fruit spotting occurred regardless of treatment, including the
uninoculated control plants that received field applications of streptomycin, indicating

pathogen spread within the field. Generally, “bird’s-eye” spotting does not affect quality

81

of processing tomatoes grown for paste but causes problems in tomatoes processed for
whole pack (21) and for tomatoes grown for the fresh market. Research on the
epidemiology of fruit spot formation is lacking, yet it is known that bird’s eye spots are a
result of flower (1 8) and young fruit infection (6). A recent study found that when
inoculated, flowers were most susceptible to infection two-days post anthesis (18). In
addition, small green fruit developed bird eye lesions when inoculated with C.
michiganensis subsp. michiganensis. Similarly, a study by Getz, et al. determined that
spots on tomato fruit caused by Pseudomonas syringae pv. tomato occurred during the
time when flowers were past anthesis and until fruit were a maximum of 3 cm in
diameter (9). Concurrent scanning electron microscopy studies revealed that the bacterial
populations inhabited trichomes prior to anthesis and populated the openings left after the
trichomes were shed indicating the source of infection sites on fruit (8).

This study verifies findings of Hausbeck, et al. (11), that to prevent severe
bacterial canker in the field, growers should initiate and sustain bactericide applications
in the greenhouse to suppress pathogen populations. Oftentimes, treatments are made in
response to disease symptoms occurring in the field. By focusing control strategies in the
greenhouse rather than in the field, growers will have better control of the pathogen and
avoid economic losses resulting from yield losses caused by this disease. Since
acibenzolar—S-methyl (13) has no direct affect on the pathogen and depends on the onset
of SAR in the plant, this product should be applied preventively. This study identifies
novel disease management strategies that upon further testing should ease reliance on

copper for control of bacterial canker.

82

LITERATURE CITED
Bay, D. M., and Wyant, D. (eds). 1999. Michigan Agricultural Statistics. MDA,
Michigan pp. 42-45.
Chang, R. J., Ries, S. M., and Pataky, J. K. 1992. Local sources of Clavibacter
michiganesis subsp. michiganensis in the development of bacterial canker on
tomatoes. Phytopathology 82:553-560.
Cole, D. L. 1999. The efficacy of acibenzolar-S-methyl, an inducer of systemic
acquired resistance, against bacterial and fungal diseases of tobacco. Crop Prot.
18:267-273.
Conover, W. J. 1980. Practical Nonparametric Statistics, 2nd ed. Wiley, New
York.
Cooksey, D. A. 1990. Genetics of bactericide resistance in plant pathogenic
bacteria. Annu. Rev. Phytopathol. 28:201-19.
Ehrig, F., and Grisbach, E. 1985. Development of bird eye spots on tomato fi'uits-
Investigations by scanning electron microscopy. Arch. Phytopathol.
Pflanzenschutz 21:419-423.
Frey, P., Smith, J. J., Albar, L., Prior, P., Saddler, G. S., Trigalet Demery, D., and
Trigalet, A. 1996. Bacteriocin typing of Burkholderia (Pseudomonas)
solanacearum race 1 of the French West Indies and correlation with genomic
variation of the pathogen. Appl. Environ. Microbiol. 62:473-479.
Getz, S., Fulbright, D. W., and Stephens, C. T. 1983. Scanning electron

microscopy of infection sites and lesion development on tomato fruit infected

83

10.

11.

12.

13.

14.

with Pseudomonas syringac pv. tomato. Phytopathology 73:39-43.

Getz, S., Stephens, C. T., and Fulbright, D. W. 1983. Influence of developmental
stage on susceptibility of tomato fruit to Pseudomonas syringae pv. tomato.
Phytopathology 73:36-38.

Hammerschmidt, R., and Becker, J. S. 1997. Acquired resistance to disease in
plants. Pages 247-289 in : Horticultural Reviews, vol. 18. Janick, J. (ed.) John
Wiley & Sons, Inc., New York.

Hausbeck, M. K., Bell, J ., Medina-Mora, C., Podolsky, R., and Fulbright, D. W.
2000. Effect of bactericides on population sizes and spread of Clavibacter
michiganensis subsp. michiganenesis on tomatoes in the greenhouse and on
disease development and crop yields in the field. Phytopathology 90:38-44.
Jones, J. B., Woltz, S. S., Jones, J. P., and Portier, K. L. 1991. Population
dynamics of Xanthomonas campestris pv. vesicatoria on tomato leaflets treated
with copper bactericides. Phytopathology 81:714-719.

Knight, S. C., Anthony, V. M., Brady, A. M., Greenland, A. J ., Heaney, S. P.,
Murray, D. C., Powell, K. A., Schultz, M. A., Spinks, C. A., Worthington, P. A.,
and, and Youle, D. 1997. Rationale and perspectives on the development of
fungicides. Annu. Rev. Phytopathol. 35:349-72.

Louws, F. J., Ritchie, D. F ., Kousik, C. S., Romero, A. M., Pollard, D. W. 1996.
Evaluation of the plant-activator CGA-245704 SOWG for control of bacterial spot
of tomato. Pages 18-19 in: 12th Annual Tomato Disease Workshop Columbus,

Ohio.

84

15.

16.

I7.

18.

19.

20.

21.

Louws, F. J ., Bell, J ., Medina Mora, C. M., Smart, C. D., Opgenorth, D.,
Ishimaru, C. A., Hausbeck, M. K., deBruijn, F. J ., and Fulbright, D. W. 1998. rep-
PCR-mediated genomic fingerprinting: a rapid and effective method to identify
Clavibacter michiganensis. Phytopathology 88:862-868.

Manley, B. S., Bassi, Jr., A. B., Tally, A. H., Young, T. R., Payan, L., Coud, G.,
and Reuss, W. 1997. Actigard (CGA 245704), a novel approach to bacterial spot
and bacterial speck control in tomato. Pages 48-49 in: 13th Annual Tomato
Disease Workshop, Indianapolis, IN.

Marco, G. M. 1983. Control of bacterial spot of pepper initiated by strains of
Xanthomonas campestris pv. vesicatoria that differ in sensitivity of copper. Plant
Dis. 67:779-781.

Medina-Mora, C. M., Hausbeck, M. K., and Fulbright, D. W. 2000. Bird's eye
lesions of tomato fruit produced by aerosol and direct application of Clavibacter
michiganensis subsp. michiganensis. Plant Dis. (Accepted pending revision).
Ricker, M. D., Emmatty, D. A., Schott, M. D., and Ozrninkowski, R. H. 1997.
Heinz tomato varieties resistant to bacterial canker. Pages 37-39 in: 10th Annual
Tomato Disease Workshop Windsor, Ontario Canada.

Ricker, M. D., and Riedel, R. M. 1993. Effect of secondary spread of Clavibacter
michiganensis subsp. michiganensis on yield of northern processing tomatoes.
Plant Dis. 771364-366.

Ricker, M. D., and Riedel., R. M. 1988. Economic importance of secondary

spread of Clavibacter michiganensis subsp. michiganensis in northem-grown

85

processing tomatoes. (Abstr.). Phytopathology 7821570.
22. Zar, J. H. 1996. Biostatistical Analysis 3rd ed. Prentice Hall, Upper Saddle River,

NJ.

86

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

llill(willmillilulllllllll