Citation
Survival of Salmonella and Shigella on Tomatoes in the Presence of the Soft Rot Pathogen, Erwinia carotovora

Material Information

Title:
Survival of Salmonella and Shigella on Tomatoes in the Presence of the Soft Rot Pathogen, Erwinia carotovora
Copyright Date:
2008

Subjects

Subjects / Keywords:
Bacteria ( jstor )
Error rates ( jstor )
Erwinia ( jstor )
Humidity ( jstor )
Information economics ( jstor )
Plant nutrition ( jstor )
Ripening ( jstor )
Salmonella ( jstor )
Shigella ( jstor )
Tomatoes ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
4/17/2006

Downloads

This item has the following downloads:


Full Text












SURVIVAL OF Salmonella AND \/nlgel// ON TOMATOES IN THE PRESENCE OF
THE SOFT ROT PATHOGEN, Erwinia Carotovora















By

JENNIFER A. JOY


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Jennifer A. Joy

































To my parents, John and Dolores Joy; their unwavering love and support made this
possible















ACKNOWLEDGMENTS

I would like to thank my supervisory committee chair (Dr. Keith R. Schneider) for

his assistance and guidance. My supervisory committee members (Dr. Douglas L.

Archer and Dr. Jerry A. Bartz) are also due many thanks for their help in my research.

Six L's Packing Company, Inc. and DiMare Fresh, Inc. are greatly appreciated for

always providing top-quality tomatoes. Without these produce suppliers, much of my

research would not have been completed. I would like to extend thanks to all of my lab

mates for all of their hard work and diligence in assisting me with research projects. This

research was funded in part by the USDA-CSREES (Grant number 00-52102-9637),

responsible for providing the .33 FTE research assistantship. Additionally, I would like

to thank the Institute of Food and Agricultural Sciences (IFAS) Statistics Department for

all of their help in analyzing my results.

Most importantly, I would like to thank my parents, John and Dolores Joy, for all

of their love and encouragement. My achievements of the past 2 years would not have

been possible without them.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

L IST O F T A B L E S ......................... ........ .... ........................ .. ........ ........... .. vii

LIST O F FIG U R E S ................................................................. ................... viii

ABSTRACT .............. ......................................... ix

CHAPTER

1 IN TRODU CTION ................................................. ...... .................

2 LITER A TU R E REV IEW ............................................................. ........................ 4

Importance of Foodborne Illness Related to Fresh Produce.......................................4
Foodbom e Illness Linked to Produce.................................... ..................................... 5
Salm onella .................................................................................................. ....... 7
Salmonellosis..................................9
,.1t.ge/ll t .............................................. ......... 10
Shigellosis ............................................. 11
E rw inia carotovora ....................................................... 12
T o m ato es ....................................................................................................... 14
Tom ato M market and R regulations ....................................................... 15
Tom atoes and Pathogens ............................................................ .... ..........16
Quorum Sensing ........................................................................................................17

3 M ATERIALS AND M ETHODS ....................................................... 20

Selection of Temperature and Relative Humidity Conditions ............... ................20
Acquisition and Maintenance of Salmonella Cultures ............................................. 21
Acquisition and Maintenance of .\l/gel/t Cultures ............................................22
Acquisition and Maintenance of Erwinia carotovora Culture ................................23
G ro w th S tu d ie s ..................................................................................................... 2 3
P reparation of Inoculum ........................................................................ ............... 24
Inoculation of Tom atoes ............................................ .... .... .......... ..... 25
Intact T om ato es ........................................................................................2 5
Shave-W wounded Tom atoes ...................................................... ...............26




v









Pathogen Recovery off Tomato Surfaces ....................................... ............... 26
Intact T om atoes ....................... .. .... ..................... .. ...... .... ...........26
Shave-Wounded Tomatoes.................................................27
Shave-Wounded Tomatoes with Combined Pathogens ....................................28

4 R E S U L T S ............................................................................................................. 3 0

Growth Levels of Salmonella, .\ lge//At and Erwinia carotovora............................ 31
Preliminary Recovery Studies: Intact Tomato Surfaces...........................................33
Recovery of Bacteria from Tomato Surfaces .................................. ............... 33
Recovery of Bacteria from Intact Tomato Surfaces ....................................... 34
Salmonella Recovery off Intact Tomato Surfaces............... ...............34
\l/ige/ll Recovery from Intact Tomato Surfaces.....................................35
Erwinia Recovery from Intact Tomato Surfaces...............................................36
Recovery of Bacteria from Shave-Wounded Tomato Surfaces ..............................38
Salmonella Recovery from Shave-Wounded Tomato Surfaces ........................38
\l/ige/ll Recovery from Shave-Wounded Tomato Surfaces.............................39
Erwinia Recovery from Shave-Wounded Tomato Surfaces ............................40
Recovery of Combined Bacteria from Shave-Wounded Tomato Surfaces ..............41
Recovery of Salmonella and Erwinia in Fall/Winter Season Conditions ..........41
Recovery of Salmonella and Erwinia in Optimum Conditions for Erwinia .......42
Recovery of Salmonella and Erwinia in Standard Ripening Room Conditions .43
Recovery of .\/ige//t and Erwinia in Fall/Winter Season Conditions ...............44
Recovery of .\l/ge//At and Erwinia under Optimum Conditions for Erwinia......46
Recovery of \l/ge//At and Erwinia in Standard Ripening Room Conditions ......47

5 D ISC U S SIO N ....................................................................... .................... .... 4 8

Recovery of Bacteria from Intact Tomato Surfaces ................................................50
Recovery of Bacteria from Wounded Tomato Surfaces.......................................53
Recovery of Combined Bacteria from Wounded Tomato Surfaces .........................55

6 C O N C L U SIO N ......... ......................................................................... ........ .. ..... .. 58

APPENDIX

A SALM ONELLA STATISTICS .............................................................................60

B SHIGELLA STA TISTICS...................................................................... ............... 70

C ERW INIA STA TISTIC S .................................................. ............................... 80

L IST O F R E FE R E N C E S ......................................... .. ............................... ................... 90

BIO GRAPH ICAL SK ETCH .................................................. ............................... 96















LIST OF TABLES


Table pge

2-1 ,l/ige/ll species designated by their serogroup and number of serotypes ..............11

2-2 Tom ato size designation................................................. .............................. 16

3-1 Temperature and relative humidity conditions selected to simulate a standard
ripening room environment (90% RH, 200C) and optimum temperature for the
growth of Erwinia at fall/winter conditions (60% RH, 270C) and a ripening
room (90% R H 27 C )............. ...................................................... ...... .... ..... 2 1

3-2 Salmonella enteritidis serovars obtained from Dr. Linda J. Harris at the
University of California, Davis: wild type serovars listed with source....................22

4-1 The logo CFU/mL reduction of individual bacteria off intact tomato surfaces
during prelim inary studies.............................................. .............................. 33

4-2 Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in
fall/winter (60% RH, 270C) conditions over three days. .......................................42

4-3 Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in
optimized conditions for Erwinia carotovora (90% RH, 270C) over three days.....43

4-4 Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in
standard ripening room conditions (90% RH, 200C) over three days....................44

4-5 Recovered .\illgel/, and Erwinia from shave-wounded tomato surfaces in
fall/winter season conditions (60% RH, 270C) over three days............................45

4-6 Recovered \l/ige/lit and Erwinia from shave-wounded tomato surfaces under
optimum conditions for Erwinia carotovora (90% RH, 270C) over three days. .....46

4-7 Recovered .\/ligetll and Erwinia from shave-wounded tomato surfaces under
standard ripening room conditions (90% RH, 200C) over three days ................ 47















LIST OF FIGURES


Figure page

4-1 Average logo CFU/mL growth curves of Salmonella serovars over a 10-hour
incubation .......................................................................... 3 1

4-2 Average logo CFU/mL growth curves of .\/ngell/ species over a 12-hour
incubation .......................................................................... 3 1

4-3 Average logo CFU/mL growth curves of Erwinia carotovora over a 21-hour
incubation ............................................................................32

4-4 Maximum average logo CFU/mL counts of Salmonella serovars, .\/ngel//
species and Erwinia (rif+) after specific incubation times..................................32

4-5 Salmonella recovery logoo CFU/mL) from intact tomato surfaces in fall/winter
(60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C) and ripening
room parameters (90% RH, 200C) over three days ............................. ...............35

4-6 ,\l/gel// recovery logoo CFU/mL) from intact tomato surfaces in fall/winter
(60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C) and ripening
room parameters (90% RH, 200C) over three days ............................. ...............36

4-7 Erwinia carotovora recovery logoo CFU/mL) from intact tomato surfaces in
fall/winter (60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C)
and ripening room parameters (90% RH, 200C) over three days.............................37

4-8 Salmonella recovery logoo CFU/mL) from shave-wounded tomato surfaces in
fall/winter (60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C)
and ripening room parameters (90% RH, 200C) over three days.............................38

4-9 ,\l/gel// recovery logoo CFU/mL) from shave-wounded tomato surfaces in
fall/winter (60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C)
and ripening room parameters (90% RH, 200C) over three days.............................40

4-10 Erwinia carotovora recovery logoo CFU/mL) from shave-wounded tomato
surfaces in fall/winter (60% RH, 270C), optimum conditions for Erwinia (90%
RH, 270C) and ripening room parameters (90% RH, 200C) over three days. .........41















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

SURVIVAL OF Salmonella AND .l/ge/la ON TOMATOES IN THE PRESENCE OF
THE SOFT ROT PATHOGEN, Erwinia Carotovora

By

JENNIFER JOY

December 2005

Chair: Keith R. Schneider
Major Department: Food Science and Human Nutrition

Recently, outbreaks of Salmonella and \,l/lg//t have been associated with the

consumption of fresh produce. Previous studies have investigated the survival of

Salmonella and .\llge/,l on fresh produce, however there is little data on the effects of

bacterial soft rot on human pathogen survival. Erwinia carotovora is the bacterial

species most commonly associated with soft rot on fresh vegetables. My study

investigated the survival of Salmonella and ./nhge//t with and without the presence of

Erwinia carotovora on intact and compromised tomato surfaces.

Whole, unripe tomatoes were artificially inoculated with either a 5-strain

Salmonella cocktail or a 2-strain .\l/lgle/l cocktail (with and without Erwinia

carotovora). Inoculated tomatoes were incubated at room temperature (-27C) under

controlled relative humidity to simulate fall/winter in Florida (60%) and optimum growth

conditions for Erwinia carotovora (90%). Additionally, inoculated tomatoes were

incubated at standard ripening-room conditions of 90% relative humidity and 200C.









All three environmental conditions allowed for the survival of Salmonella on

intact tomato surfaces. Significant population decreases were observed, but viable cells

were still present after 3 days. Optimum growth conditions for Erwinia carotovora

allowed the best survival of.\/ige//At and Erwinia on intact tomato surfaces. Wounded

tomato surfaces allowed growth of the bacteria except for .\/ige/l/ at fall/winter season

conditions. No viable cells were recovered over the span of 3 days. Salmonella and

,/lig/llt on wounded surfaces reached peak growth at standard ripening room conditions

(90% RH, 200C), while Erwinia optimum conditions for growth were 90% RH, 27C. In

combined bacteria studies on wounded tomato surfaces using Salmonella and Erwinia,

peak growth of Salmonella occurred under optimum conditions for growth of Erwinia

(4.81 logo CFU/mL), while peak growth of Erwinia occurred under standard ripening-

room conditions (6.26 logo CFU/mL). For studies combining .l/Vge/lit and Erwinia,

peak growth occurred for both microorganisms when held under optimum growth

conditions for Erwinia (4.84 logo CFU/mL and 6.55 logo CFU/mL respectively).

My results suggest that temperature and relative humidity (not temperature alone)

play an important role in the growth and survival of bacteria on both intact and wounded

tomato surfaces. The risk of foodborne illness from fresh produce can be lessened by

ensuring good culling practices and establishing a lower humidity in packinghouses to

prevent the growth of bacteria on produce, specifically tomatoes.














CHAPTER 1
INTRODUCTION

In recent years, increases in the number of outbreaks of human illness have been

associated with consumption of fresh fruits and vegetables (Tauxe et al. 1997). Several

factors may have contributed to this increase including the emergence of new pathogens,

adaptation of pathogens to environmental stresses imposed by fruits and vegetables,

increased global travel facilitating exposure to pathogens from other geographical areas,

and agricultural practices that use improperly treated irrigation water or animal manure

(Beuchat 1995; Tauxe et al. 1997; Wade and Beuchat 2003).

Salmonella serotypes cause approximately 1.5 million cases of foodborne

gastroenteritis annually in the U.S., resulting in approximately 15,000 hospitalizations

and 500 deaths (Mead et al. 1999). .\/ige/ll spp. have a relatively low infectious dose,

requiring as few as 10 to 1000 cells to cause illness (Smith 1987). Yearly in the U.S.,

.\/ige/l spp. are responsible for approximately 450,000 cases of foodborne illness,

resulting in 1,300 hospitalizations and 14 deaths (Mead et al. 1999). The Centers for

Disease Control and Prevention (CDC) report that the number of produce-associated

outbreaks and cases of foodborne illness has more than doubled between the periods of

1973 to 1987 and 1988 to 1991 (Tauxe et al. 1997). In 1999, the FDA conducted a

survey of produce commonly consumed in the U.S., including tomatoes, cantaloupe,

green onions, strawberries, and celery. They tested a total of 1003 samples for the

presence of Salmonella, .l/Vge/Ii, or E. coli 0157:H7. Of the samples analyzed, 4% were

found to be contaminated with either Salmonella or ,\lige/lat (FDA 2001). In response to









that study, the FDA repeated this produce survey in 2000 and determined that 1% of the

samples was contaminated with either Salmonella or .\/nge//t and none were found to be

contaminated with E. coli 0157:H7 (FDA 2003).

High-acid fruits and vegetables (pH less than 4.6), previously were considered

unable to support the growth of bacteria that cause human infections (Bagamboula et al.

2002). Some foodborne pathogens have shown surprising tolerance to stressful

environments. Salmonella serovars have been shown to grow at temperatures ranging

from 20C (Baker et al. 1986) to 540C (Droffner and Yamamoto 1992) and over a pH

range from 3.99 (Asplund and Nurmi 1991) to 9.5 (Holley and Proulx 1986). Several

recent outbreaks of salmonellosis in the U.S. have been associated with consumption of

uncooked tomatoes, including S. enterica serotypes Javiana and Montevideo (Hedberg et

al. 1999). Additionally, raw tomatoes have been implicated in outbreaks involving S.

enterica serotype Baildon (Cummings et al. 2001). The reported occurrence of

foodborne shigellosis, primarily caused by .\/nge/lt sonnei and ./n/ge/ltflexneri, is lower

than that of salmonellosis or other enteric pathogens. Each year there are a significant

number of outbreaks of shigellosis, many associated with consumption of fresh produce

(Smith 1987; Mead et al. 1999). The most important contributing factor leading to

/nlge//t foodborne outbreaks is poor personal hygiene of a food handler (Smith 1987).

Fruits and their products should be handled properly to prevent contamination. The

increase in worldwide food trade and fresh produce consumption, and the popularity of

minimally processed fruits and vegetables may have a significant role in introducing

/nge//At spp. to these types of foods (Bagamboula et al. 2002).









Most of the bacteria responsible for spoilage of fresh-cut fruits and vegetables are

Gram-negative. Erwinia carotovora is the species most commonly associated with decay

of vegetables (Lund 1983) such as potato, tomato, onion, corn, rice, and sugar beets

(Aysan et al. 2003). Soft rot Erwinias cause tissue maceration (referred to as soft rot

disease), by producing cell-wall-degrading enzymes (Toth et al. 2003). The main

weapon used by soft rot Erwinia is coordinated production of high levels of multiple

exoenzymes including pectinases and proteases, which break down plant cell walls and

release nutrients for bacterial growth (Barras et al. 1994).

My study evaluated the survival and recovery of Salmonella spp. and .\/gell,// spp.

from tomato surfaces individually and in the presence of the soft-rot plant pathogen,

Erwinia carotovora. Intact and shave-wounded tomatoes were inoculated with known

concentrations of the rifampicin-resistant variants of the bacteria and combinations of

Salmonella-Erwinia and .\lge/ll, -Erwinia. Bacterial recovery off of intact tomato

surfaces was accomplished by a rub-shaking method (Burnett and Beuchat 2001; Harris

et al. 2002; Zhuang et al. 1995), while the shave-wounded tomato samples were excised

and stomached. Tomatoes were kept in an environmental humidity chamber for 3 days to

maintain specific temperature and humidity conditions. Conditions were established to

simulate the typical Florida fall/winter production season (60% RH, 27C), optimum

conditions for the survival ofErwinia carotovora (90% RH, 27C), and standard tomato

ripening rooms (90% RH, 200C).














CHAPTER 2
LITERATURE REVIEW

Importance of Foodborne Illness Related to Fresh Produce

During the last 25 years, the number of foodborne disease outbreaks and illnesses

has increased (Mead et al. 1999; Beuchat 2002). Outbreaks of foodborne illness have

been attributed to the consumption of contaminated fresh vegetables and, to a lesser

extent, fruits (Beuchat 1995; Brackett 1999). Despite great advances in science and

technology, foodborne illness is still a major problem even in highly developed countries

such as the U.S. and Canada. Additionally, there has been a significant increase in the

different types of foods that have been found to harbor illness-causing foodborne bacteria

(Samelis and Sofos 2003).

Consumers recognize that fresh fruits and vegetables are important components of

a healthy and balanced diet. In recent years, increased demand for fresh fruits and

vegetables has arisen from studies showing positive correlations between diet and health.

Additionally, whole and lightly processed fruits and vegetables have gained interest due

to greater selection and year-round availability (Lucier et al. 2000). Contamination of

fruits and vegetables by pesticides, plant and human pathogens, and other pollutants is

becoming a larger concern to consumers and to the fresh-produce industry (Beuchat

1995; Brackett 1999; Wilson and Droby 2001).

Salmonellosis outbreaks have been attributed to contaminated tomatoes, bean

sprouts, cantaloupe, and watermelon (Beuchat 1996); and have been known to cause

invasive disease or reactive arthritis (Altekruse et al. 1997). A .\lige/lltflexneri









gastroenteritis outbreak was associated with green onions (Beuchat 1996) and an

outbreak of E. coli 0157:H7, the enterohemorrhagic serotype, has been linked to the

consumption of contaminated cantaloupes (Beuchat 1996).

Foodborne Illness Linked to Produce

Incidence of foodborne illness attributed to fresh produce is low, but there is

always the possibility of fruits and vegetables being contaminated. Concern about

contaminated fruits and vegetables is well founded because of the increase in

consumption of fresh produce and the increase of produce imported from countries where

standards of sanitation and handling are more lenient than here in the U.S. (Beuchat

1996). Salmonella serotypes cause approximately 1.5 million cases of foodborne

gastroenteritis annually in the U.S. resulting in approximately 15,000 hospitalizations and

500 deaths (Mead et al. 1999).

Wells and Butterfield (1997) showed a strong statistical association between the

occurrence of bacterial soft-rot plant pathogen (Erwinia carotovora) and the presence of

Salmonella on tomatoes. Of the bacterial isolates recovered from the tomatoes used in

the study, 30% were confirmed to be Salmonella. Liao and Sapers (1999) found an

antagonistic relationship between soft-rotting bacteria and L. monocytogenes on potato

slices. Additionally, Pseudomonasfluorescens prohibited the growth of Listeria, but E.

carotovora enhanced growth of this human pathogen on potato slices. It is unclear

whether these responses are caused by positive or negative interactions between the

pathogen and the spoilage organisms or these responses are caused by increased

availability of nutrients in the macerated tissue produced as a result of soft rot (Novak et

al. 2003).









To determine the contamination rate of imported fresh produce, in March 1999, the

U.S. Food and Drug Administration (FDA) conducted a survey of 1003 samples from 21

countries. The samples included broccoli, cantaloupe, celery, cilantro, loose-leaf lettuce,

parsley, scallions, strawberries and tomatoes. Samples were analyzed for Salmonella,

.\li/ge/ll and E. coli 0157:H7. Four percent (44 of 1003 samples) were contaminated

with either Salmonella or .\/nge//t Of these contaminated samples, 80% (35 of 44

samples) were contaminated with Salmonella, and 20% (9 of 44 samples) were

contaminated with .\l/ge/ll The produce with the most contamination of Salmonella and

,/i5g/llt was cantaloupe and cilantro (FDA 2001).

As a follow-up to these findings, in March of 2000, the FDA conducted another

produce survey including 1000 samples from approximately 18 states. High volume

produce such as cantaloupe, celery, cilantro, loose-leaf lettuce, parsley, scallions,

strawberries and tomatoes were collected and analyzed for contamination with

Salmonella, .\/nge/ll and E. coli 0157:H7. Of the total 1028 domestic samples, 99%

were not contaminated with any of the mentioned pathogens. Eleven samples (1% of

total samples) were contaminated with either Salmonella or .\l/ge/lt and no samples

analyzed were contaminated with E. coli 0157:H7. Of the 11 contaminated samples, six

(55%) were contaminated with Salmonella and five (45%) were contaminated with

1/ge/lli, with cantaloupe and scallions resulting in the majority of positive samples

(FDA 2003).

Often fresh produce is washed in tap water and left to air dry. Kelman (1956)

found that the viability of a soil-borne bacterium, Ralstonia solanacearum could survive

in tap water for several months. In response to this research, a study by Liao and









Shollenberger (2003) was performed to determine the suitability of using sterile water

and phosphate buffered saline for preservation of bacteria that were pathogenic to plants

or humans. Their findings show that soft rot bacteria including Erwinia spp. remain

viable for 12-16 years in sterile distilled water, while the foodborne pathogen Salmonella

was able to survive for 5 years (Liao and Shollenberger 2003).

Fresh produce is also exposed to sanitizers in the processing water, primarily

hypochlorite. Many consumers incorrectly assume that the sanitizers are used to treat the

fruit or vegetable, when in actuality, the sanitizers are primarily used to maintain

bacteriological quality of the water (Bracket 1999). In a study by Zhuang et al. (1995),

the effect of sanitizers and surfactants were of minimal value in reducing populations of

Salmonella in tomato fruit demonstrating that sanitizers may help but cannot guarantee

the complete elimination of pathogens from produce (Brackett 1999).

Salmonella

The foodborne organism Salmonella was first recognized in the late 1800's. In

1885, veterinary pathologist D.E. Salmon isolated the microorganism Bacillus cholerae-

suis from pigs suffering from hog cholera (Cox 2000a). Following this discovery, other

similar microorganisms were isolated from foodborne outbreaks originating from food

animals. To accommodate these new findings the genus Salmonella was created in honor

of Salmon. The genus Salmonella is a member of the family Enterobacteriaceae and is

comprised of facultative anaerobic, oxidase-negative, catalase positive, Gram-negative,

rod-shaped bacteria, ranging in size from 0.7-1.5 x 2-5 [m (Cox 2000a). Most

Salmonella strains are motile and are capable of fermenting glucose with the production

of acid and gas (FDA 2005).









Salmonella spp. can be separated into serotypes or serovars by characterizing two

major antibodies designated the O (somatic antigen) or H (flagellin antigen). The O

antigen is divided into serogroups, based on differences in epitopes of the major outer

membrane component of Gram-negative bacteria, the lipopolysaccharide (LPS). The H

antigen is used to establish variations associated with the subunit proteins of the flagella

(Cox 2000a).

There are several environmental conditions that affect the growth, death or survival

of Salmonella including temperature, pH and water activity (aw). The growth range for

this organism is 20C (Baker et al. 1986) to 540C (Droffner and Yamamoto 1992), with

the optimum being 370C (Cox 2000a). This optimum temperature is to be expected given

the natural environment for most Salmonella strains of public health significance are

found in the gastrointestinal tract of warm-blooded animals. Often foods high in solid

content, especially protein or fat, can offer protection for the microorganism at high

temperatures (Cox 2000a).

Salmonella are capable of growing on defined media without special growth

factors, with most strains being aerogenic (Buchanan and Gibbons 1974). The pH

optimum range for growth of Salmonella is 6.5-7.5 and survival is possible up to 9.5

(Holley and Proulx 1986) and as low as 3.99 (Asplund and Nurmi 1991). An increase in

temperature has been shown to increase sensitivity to low pH, as well as the presence of

food additives like salt or nitrite. The optimum water activity for this microorganism is

0.995, with no growth being observed below 0.93. The survival time of the bacteria

increases as the water activity decreases, allowing survival in low-moisture foods such as

peanut butter and chocolate for several months (Cox 2000a).









Salmonellosis

After its discovery in 1885, Salmonella was recognized as an agent of disease and

is still a major threat to animals and humans. Salmonella serotypes cause an estimated

1.5 million cases of foodborne gastroenteritis annually in the U.S. Of these cases, 15,000

hospitalizations and 500 deaths occur (Mead et al. 1999). Most people who are infected

with Salmonella develop diarrhea, fever, and abdominal cramps 12 to 72 hours after

infection. The illness lasts approximately 4 to 7 days, and most people recover without

treatment. In some people the diarrhea may be so severe that the patient needs to be

hospitalized. In these patients, the Salmonella infection may spread from the intestines to

the blood stream, and then to other body sites and can cause death unless the person is

treated promptly with antibiotics. The elderly, infants, and those with impaired immune

systems are more likely to have a severe illness (CDC 2004a).

The food products usually responsible for causing illness are eggs and poultry

meat. Fresh fruits and vegetables have also been implicated in foodborne outbreaks

(Yoon et al. 2004). Most recent cases of Salmonella associated with fresh fruits and

vegetables are suspected to result from the improper storage or handling of prepared

foods that initially carried the bacteria as surface contamination. Salmonella that are

present on fruits and vegetables are capable of multiplying if specific extrinsic factors are

there, such as improper refrigeration during storage and preparation, poor product quality,

or the presence of bacterial soft rot (Wells and Butterfield 1999). Studies have shown

that tomatoes have been contaminated with S. Javiana (CDC 2002), S. Montevideo (CDC

1993, Hedberg et al. 1994), S. Poona (CDC 1991) and S. Baildon (Cummings et al.

2001). S. Oranienberg (CDC 1979) and S. Javiana (Blostein 1991) have both been

implicated in infections due to the consumption of precut watermelon, while S. Chester









(Ries et al. 1990) and S. Poona (CDC 1991) have been found to be the cause of

foodborne illness outbreaks in cantaloupe. Salmonellosis can lead to invasive disease or

reactive arthritis (Altekruse et al. 1997).

Shigella

,\hlig/lt was named after the Japanese bacteriologist, K. Shiga, who was first to

discover the dysentery bacillus (Buchanan and Gibbons 1974). .\/nge//t is a Gram-

negative, non-motile, non-sporeforming, rod-shaped bacteria (Lampel et al. 2000).

.\/nge//t is a facultative anaerobe that is capable of growing at temperatures as low as 6C

and as high as 480C. The pH range for this organism is 4.8-9.3 and can survive 5-10 days

in acidic foods such as orange juice and as much as 50 days in foods with a more neutral

pH such as milk, flour and eggs. This bacterium is a member of the family

Enterobacteriaceae and is closely related to Escherichia coli (Lampel et al. 2000). The

species S. dysenteriae, S. flexneri, S. boydii and S. sonnei make up the genus .\lVgel//A and

are all pathogenic to humans (Beuchat 1996). These four species have different virulence

levels with Strain D, S. sonnei, being the causative agent of most cases of .\/ige//A-related

diarrhea shigellosiss) (Lampel et al. 2000). The geographic distributions and

epidemiology for these species is different. S. dysenteriae, mainly found in the Indian

subcontinent, is responsible for the most severe epidemics of dysentery. S. flexneri and S.

sonnei are most commonly isolated in developed countries while S. boydii is rarely

isolated in developed countries (Lampel et al. 2000). Humans harbor .\/lge//t and the

organism is easily spread from host to host. The source of infection is usually food or

water that has been contaminated by the feces of human carriers (ICMSF 2002).









.,/liellt can be classified into four groups; A, B, C, and D, based on biochemical

reactions and the O antigen type. These groups are then subdivided into serotypes based

on the O antigen (Table 2-1).

Table 2-1. ./nge//At species designated by their serogroup and number of serotypes
Species Serogroup Serotypes
S. dysenteriae A 15
S. flexneri B 6
S. boydii C 19
S. sonnei D 1

It is estimated that 450,000 cases of shigellosis occur every year in the U.S. (Mead

et al. 1999). Shigellosis is mainly a disease found in developing countries where sanitary

waste treatment and water purification are not effective. .\/lge/lt has a low infectious

dose, which allows for high rates of communicability between humans and in some cases

subhuman primates. It has been determined that less than 100 cells of .\/ige//t are

capable of producing an illness (DuPont 1990). Consequently, .\/nge/lt is capable of

easily infecting a crowded population most frequently by the fecal-oral route. .\/lge/ll is

commonly found in water polluted with human feces and can be spread by flies, food and

among people with unacceptable personal hygiene (Smith 1987). After being infected

with .\l/gelil, the incubation period is between 1 and 7 days with symptoms first

appearing after 3 days. It is possible to be a carrier of this microorganism for up to four

weeks (Lampel et al. 2000).

Shigellosis

\/nlge//It species, mainly S. sonnei and S. flexneri are the cause of approximately

450,000 cases of foodborne illness in the U.S. yearly. Of these, 1,300 result in

hospitalizations and 14 deaths (Mead et al. 1999). \/nge//At is considered an invasive

pathogen and shigellosis occurs when virulent ,\/ge//At organisms attach to, and penetrate









epithelial cells of the large intestine. Once they have invaded the tissue, the bacteria are

released into the cytoplasm of the cell and multiply intracellularly. This infection is

spread to the neighboring cells through projections from the originally infected cell

(Lampel et al. 2000). The intestinal epithelial cells being attacked continue this cycle

causing damage resulting in dysentery. These virulence factors are present on a large 220

kb virulence plasmid. The virulence proteins contained here are regulated by different

environmental stimuli, with temperature being most influential. Virulence gene

expression in .\/nge//l is repressed at 300C and enabled at 370C (Lampel et al. 2000).

A large S. sonnei outbreak involving 347 people was traced to commercially

distributed shredded lettuce. It was determined that an infected handler was

contaminating the lettuce during the shredding process (Brackett 1992). More recently,

two midwestern U.S. outbreaks of S. flexneri have been associated with green onions,

resulting in gastroenteritis (Beuchat 1996).

Erwinia carotovora

The bacterial plant pathogen Erwinia is known to cause soft rot in many crops

worldwide. The predominant plant pathogens in this genus consist of Erwinia

carotovora ssp. atroseptica (Eca), E. carotovora ssp. carotovora (Ecc) and E.

(c hl)% (,thelni (Ech). The soft-rot Erwinia are members of the Enterobacteriaceae, along

with other plant pathogens such as Erwinia amylovora and human pathogens such as E.

coli, .\hge//lt spp., Salmonella spp. and Yersinia spp (Toth et al. 2003).

Ech is known to cause stem rot on tomatoes in greenhouses. It is capable of

inducing wilt throughout the whole plant, which results in the entire plant collapsing.

The disease first affects the roots of plants, most likely by seed-borne or soil-borne









infection and then can spread to other plants by cultural practices. When conditions for

disease are favorable, the symptoms can develop quickly (Aysan et al. 2003).

Ecc mainly affects crops in subtropical and temperate regions and has probably the

widest host range, including brussel sprouts, carrots, celery, cucumbers, turnips, and

potatoes, although there are many other crops that are rotted by these microorganisms

post-harvest (Toth et al. 2003). A pathogenic isolate of Ecc was injected into the centers

of healthy cucumber fruits attached to the vine without causing disease. However, the

bacterium was detected in the internal tissues of fruits harvested from the inoculated

plants (Guo et al. 2001).

The soft rot Erwinia are found on plant surfaces and in soil where they may enter

the plant through wound sites or natural openings on the plant surface. Once inside the

plant the bacteria survive in the vascular tissue and intercellular spaces where they

remain inactive until environmental conditions become suitable for disease development

(Perombelon and Kelman 1980; Perombelon and Salmond 1995). In addition to free

water and oxygen depletion, temperature is an important factor in disease development,

and can influence which of the soft rot Erwinia cause disease. For example, it was shown

that a soil temperature of 200C was an important transition point, above which Eca, and

below which Ech, were not apparently pathogenic (Perombelon et al. 1987a). In a study

by Perombelon et al. (1987b), the abilities of the soft rot Erwinia to grow at different

temperatures were clearly displayed in vitro, where it was used to differentiate the

pathogens. The findings showed that at 270C all three pathogens grew; at 33.50C, only

Ecc and Ech grew; at 370C only Ech grew (Perombelon et al. 1987b).









In addition to differences in growth, a thermal regulation of the production of cell

wall degrading enzymes or exoenzymes, has been demonstrated (Nguyen et al. 2002).

The main weapon employed by the soft rot Erwinia is the coordinated production of high

levels of multiple exoenzymes, including pectinases and proteases, which break down

plant cell walls and release nutrients for bacterial growth (Barras et al. 1994; Perombelon

2002; Thomson et al. 1999). Endoglucanase activity is capable of breaking down

cellulose in the primary and secondary cell walls of the host plant. Pectinases are the

main exoenzymes involved in disease development. These exoenzymes break down and

utilize pectins in the middle lamella and plant cell walls, causing tissue collapse, cell

damage and cell leakage (Toth et al. 2003).

Tomatoes

Lycopersicon esculentum, commonly referred to as the tomato, is a member of the

Solanaceae family (Olson et al. 2004). Tomatoes are mostly water averaging 93.76

grams per 100 grams of edible fruit. Tomatoes are high in the minerals magnesium,

potassium and phosphorus. Tomatoes are also good sources of lycopene, vitamin C,

vitamin A and folate (USDA 2004). One medium sized tomato provides 40% of the

RDA of vitamin C (ascorbic acid), 20% of the RDA of vitamin A, significant amounts of

potassium, dietary fiber, calcium, with as few as 35 calories (Sargent 1998).

Ripe tomatoes are soft, bruise easily, and begin to decline in quality after only a

few days. Tomatoes ripen off the vine in response to the natural ripening chemical

ethylene, which is produced by the fruit (Sargent 1998). Traditionally, growers pick the

fruits in the green-mature stage just as the fruit reach full size. The fruit is then shipped

to other locations and can resist bruising or rotting because of its firmness. The fruits are









usually red by the time they reach their destination, or they can be induced to ripen with

the application of an ethylene gas (CDC 2004b).

Tomatoes are currently one of the most popular vegetables among American

consumers, with Florida producing 40% of all commercially grown fresh tomatoes

(Sargent 1998). Tomatoes are members of the fruit family but are served and prepared as

a vegetable. The National Cancer Institute has shown through extensive research that a

diet rich in tomato-based foods can decrease the incidence of prostate cancer (CDC

2004b).

Tomato fruits have a thin epidermis which makes them easily compromised by

mechanical pressure, which can result in punctures, cracks, abrasions, and insect wounds

that render the fruit susceptible to preharvest and postharvest microbial invasion. The

stem scar tissue is also capable of absorbing water and any microorganisms that may be

present (Bartz and Showalter 1981). Tomatoes should be stored at cool room

temperature since storage below about 160C can damage the fruit and lead to poor quality

(Sargent 1998). They should appear red or reddish-orange when ripe and should be free

from bruises, blemishes or cracks (CDC 2004b).

Tomato Market and Regulations

The U.S. is one of the five leading tomato-producing countries. In 1985, the per

capital consumption of raw tomato fruit in America was 16.6 lb, which in 1995 increased

to 18.8 lb (Guo et al. 2001). This increase in consumption is due to the expansion of the

domestic greenhouse industry. Tomatoes are second only to potatoes in both U.S. farm

value and vegetable consumption. Over the past 20 years, U.S. annual per capital use of

tomatoes and tomato products has increased nearly 30 percent, making a farm value of









$1.8 billion (Lucier et al. 2000). The U.S. Standards for Grades of Fresh Tomatoes give

size designations (Table 2-2):

Table 2-2. Tomato size designation
Size Designation Minimum Diameter (inches) Maximum Diameter (inches)
Small 24/32 29/32
Medium 28/32 217/32
Large 216/32 225/32
Extra Large 224/32

When grading tomatoes, the following color terms may be used:

* Green: The color of the tomato is completely green in color.

* Breakers: A definite break in color from green to yellow or orange on not more
than 10% of the surface.

* Turning: More than 10% but less than 30% in the color change from green to red.

* Pink: with more than 30% but less than 60% of the surface shows a pink or red
color.

* Light red: More than 60% but less than 90% of the surface is a red color.

* Red: More than 90% of the surface being red in color (USDA 1997).

Tomatoes and Pathogens

Tomatoes have been involved in many multistate outbreaks of Salmonella (CDC

2002, Cummings et al. 2001, Hedberg et al. 1999). Several of these outbreaks can be

traced back to the packinghouse where tomatoes are normally dumped into a common

water bath (Hedberg et al. 1999). If the wash water or surface of the tomatoes is

contaminated, bacteria can possibly contaminate the interior of the fruit resulting in faster

decay. It is also possible for bacteria to survive on the fruit's surface and then be

transferred to the flesh during handling or cutting (Ibarra-Sanchez et al. 2004).

Field-grown plants are continuously exposed to many soil-inhabiting

microorganisms. Root diseases are often caused by interactions between organisms.









Plant parasitic nematodes are capable of making plant roots more susceptible to invasion

and pathogenesis by other microorganisms. In a recent study by Beuchat et al. (2003),

the potential role of an endoparasitic root knot nematode, Meloidogne incognita, in

facilitating the entry of Salmonella into tomato root tissues was studied. It was

determined that 31 of 36 samples infested with M incognita and Salmonella were

positive for Salmonella four weeks after inoculation.

In a study by Guo et al. (2001), the possibility of internalization of Salmonella in

tomato fruits developed from inoculated flowers and stems was observed. Salmonella

was detected in stem scar tissue and pulp of tomatoes from inoculated plants. It was also

detected on or in tomatoes from plants receiving stem inoculation before or after flower

set, and on or in tomatoes that developed from inoculated flowers. The highest

percentage of Salmonella was found on the surface of the tomato and around stem scar

tissue (Guo et al. 2001).

Quorum Sensing

Quorum sensing is understood to occur as a cell density dependent signaling

system that occurs in many genera of bacteria (Smith et al. 2003). Initiation of a

concerted population response depends on the population reaching a minimal population

unit or "quorum" (Fray 2001). This mechanism enables bacteria to alter several cellular

functions including sporulation, biofilm formation, bacteriocin production, and virulence

responses among others. Quorum sensing involves cell-to-cell communication and is

controlled by extracellular chemical signals referred to as autoinducers or bacterial

pheromones, produced by the bacteria when certain cell densities are reached. When

these specific cell densities or signals are reached, target genes are either repressed or

activated once recognized (Smith et al. 2003).









An advantage of quorum sensing is that it enables bacteria to access favorable

environments or nutrients. It also increases bacteria's ability to defend themselves

against eukaryotic hosts, competing bacteria, or environmental stresses. Cells are also

able to differentiate into morphological forms better adapted to survive in hostile

environment (Smith et al. 2003).

A common autoinducer in gram negative bacteria, the N-acyl homoserine lactones

(AHLs) is produced when cell density increases, and is critical in regulating genes

important for dissemination and virulence in animal and plant pathogens (Molina 2003).

AHLs are one of the most widespread and best understood families of signal molecules

among Gram-negative bacteria. These molecules can vary greatly with the presence or

absence of an acyl chain C3 substituent (oxo- or hydroxy-) and length of the N-acyl side

chain (four to 14 carbons) (Fray 2001). AHL regulated phenotypes are capable of

facilitating interactions between the producing organism and the surrounding

environment, including both pathogenic and symbiotic relationships with higher

organisms (Manefield et al. 2001).

AHLs are known to regulate the diverse enzymes and toxins produced by E.

carotovora. Carl ofE. carotovora produces 3-oxo-C6-HSL, which is responsible for the

induction of the secreted plant cell wall-degrading exoenzymes and of the antibiotic

carbapenem (Fray 2001). One method used to disrupt quorum sensing in E. carotovora is

the introduction of the aiiA gene cloned from Bacillus sp. into transgenic tobacco and

potato plants (Molina 2003). In a tobacco test system, E. carotovora carl mutants are

completely avirulent. They are incapable of macerating plant tissue or multiplying in

plant since they lack pectin lyase, pectate lyase, polygalacturonase, cellulase and









protease (Fray 2001). Once the gene was expressed and AHL-lactonase were produced,

the quorum sensing systems were paralyzed which resulted in increased plant disease

resistance (Molina 2003).

Another method involves constructing transgenic tobacco lines that express the E.

carotovora AHL gene, expl. Ectopic production of bacterial AHL by the transgenic

plants tricks the pathogen into prematurely secreting virulence factors, such as

pectinolytic enzymes, when cell populations are insufficient for infection. This is thought

to trigger host plant defenses resulting in the observed disease resistance. Since

genetically modified crops are not accepted in many countries, a better strategy would be

the application of microorganisms with natural ability to degrade AHLs (Molina 2003).














CHAPTER 3
MATERIALS AND METHODS

Whole and injured tomatoes were used. Tomato surfaces and shave-wounded

blossom scars were inoculated with several bacterial cocktails. These included a

Salmonella cocktail comprised of five rifampicin-resistant serovars, a \/nlge//t cocktail

comprised of two rifampicin-resistant species and a serovar of Erwinia carotovora

resistant to rifampicin. There were several sets of inoculated tomatoes including an

Erwinia and Salmonella cocktail combination, Erwinia and .lnge/All cocktail

combination, Salmonella cocktail alone, \/nge//At cocktail alone, and Erwinia alone.

There was also an uninoculated group of tomatoes to serve as the control. Recovery of

the pathogens from tomato surfaces and shave-wounded blossom scars was monitored on

Days 0, 1, 2, and 3. The tomatoes were maintained at three different temperature and

relative humidity conditions to simulate various storage conditions.

Selection of Temperature and Relative Humidity Conditions

The Florida Automated Weather Network (FAWN) (University of Florida Institute

of Food and Agricultural Sciences 2003) weather archives allowed for the selection of

temperature and relative humidity settings to represent the Florida fall/winter tomato

production season and ripening room conditions (Table 3-1). To determine the effects of

humidity on the growth and survival of Erwinia, the temperature was kept at an optimum

with varying humidity to represent standard ripening rooms and Florida fall/winter

tomato production season. The chosen parameters were used to simulate an open-air

packinghouse environment.









Table 3-1. Temperature and relative humidity conditions selected to simulate a standard
ripening room environment (90% RH, 200C) and optimum temperature for the
growth of Erwinia at fall/winter conditions (60% RH, 270C) and a ripening
room (90% RH, 270C)
Simulated environment Relative humidity (%) Temperature (OC)
Standard tomato ripening room 90 20
Erwinia optimum temperature with 90 27
tomato ripening room humidity
Erwinia optimum temperature with 60 27
Florida fall/winter production season
humidity

Acquisition and Maintenance of Salmonella Cultures

Salmonella serovars were received from Dr. Linda J. Harris (University of

California, Davis, Department of Food Science and Technology). My study used five

Salmonella enteritidis serovars; Agona, Gaminara, Michigan, Montevideo, and Poona

(Table 3-2). The serovars obtained were adapted to rifampicin (rif+) at the University of

Florida using methods described by Lindeman and Suslow (1987). The five Salmonella

serovars (rif+) were transferred to PROTECT TM Bacterial Preservers (Scientific Device

Laboratories, Des Plaines, IL) and stored at -700C. The serovars were also transferred to

NA (rif+) slants and stored at 40C.

Rifampicin is an antibiotic capable of inhibiting protein synthesis of mammalian

cells and it is freely soluble in methanol (Merck Index 2001). A 10,000 ppm (1%) stock

solution of rifampicin was utilized throughout this study. The stock solution was

prepared by dissolving 1.0 g of rifampicin (ICN Biomedical, Inc., Aurora, OH) dissolved

in 100 mL of high performance liquid chromatography (HPLC) grade methanol (Fisher

Scientific International, Fair Lawn, NJ). The stock solution was filter sterilized,

protected from light, and stored at room temperature. The Salmonella serovars were

adapted to 200 ppm rifampicin (rif+) initially and then lowered to 80 ppm rifampicin









(rif+) during experimentation to decrease the amount of stress placed on the pathogen

while still maintaining selectivity. Nutrient AgarTM (NA) (BD, Sparks, MD),

supplemented with 80 [tg/mL rifampicin (rif+) antibiotic was used to recover Salmonella

from inoculated surfaces. The rifampicin marker allowed differentiation of inoculated

Salmonella from naturally present bacteria that may have been present on the samples

(Burnett and Beuchat 2001; Lukasik et al. 2001).

Table 3-2. Salmonella enteritidis serovars obtained from Dr. Linda J. Harris at the
University of California, Davis: wild type serovars listed with source
Serovar Designation Serovar Name Origin
LJH517 Agona Alfalfa sprouts
LJH518 Gaminara Orange juice
LJH521 Michigan Cantaloupe
LJH519 Montevideo Human isolate from tomato outbreak
LJH630 Poona Human isolate from tomato outbreak

Acquisition and Maintenance of Shigella Cultures

'.\lge/Aflllexneri (LJH607) was obtained from Dr. Linda Harris (University of

California, Davis, Department of Food Science and Technology). l.\/ge//l sonnei

(ATCC 9290) was purchased from the American Type Culture Collection (ATCC,

Manassas, VA). The serovars obtained were adapted to rifampicin (rif+) at the

University of Florida using methods described by Lindeman and Suslow (1987). The

two \l1/ge//l serovars (rif+) were transferred to PROTECT TM Bacterial Preservers and

stored at -700C. The .\/nge/lt species were also transferred to NA (rif+) slants and stored

at 40C. The rifampicin was prepared using the same method as stated above. The

\l/n /llt serovars were adapted to 200 ppm rifampicin (rif+) and then reduced to 80 ppm

during experimentation to minimize stress on the cells while maintaining selectivity.

Nutrient Agar (NA) supplemented with 80[tg/mL rifampicin (rif+) antibiotic was used to

recover .\l/ge//At from inoculated surfaces. The use of rifampicin resistant .\lnge/lt









allowed for differentiation from background bacteria and only allow for growth of those

pathogens adapted to rifampicin (rif+).

Acquisition and Maintenance of Erwinia carotovora Culture

Erwinia carotovora subsp. carotovora (SR38) was obtained from Dr. Jerry A.

Bartz (University of Florida, Department of Plant Pathology). This strain of Erwinia

carotovora (SR38) was obtained from a shipment that was rejected due to decay. This

serovar was adapted rifampicin (rif+) at the University of Florida using methods

described by Kaspar and Tamplin (1993). The serovar was then transferred to

PROTECTTM Bacterial Preservers and stored at -700C. Erwinia carotovora was also

transferred to NA (rif+) slants and stored at 40C. The Erwinia were adapted to a

maximum concentration of 200 ppm (rif+) and then lowered to 80 ppm during

experimentation to allow for growth while still being a selective medium. Nutrient Agar

(NA) supplemented with 801g/mL rifampicin (rif+) antibiotic was used to recover

Erwinia from inoculated surfaces. Rifampicin (rif+) was used to eliminate any

background bacteria that may have been present on the samples.

Growth Studies

Growth studies were performed in triplicate for each pathogen and its respective

serovars to determine the rate of growth for each. These studies were done to ensure that

each cocktail contains the same quantity of each serovar (CFU/mL), preventing

domination of one serovar over the others. The five Salmonella serovars (rif+), two

\1/nlge/t serovars (rif+) and the Erwinia carotovora serovar (rif+) were revived off

PROTECT TM Bacterial Preservers by aseptically transferring one bacterial preserver into

10 mL of Tryptic Soy Broth (TSB) (BD, Sparks, MD) supplemented with 80 [l of

rifampicin. The Salmonella and .ltlge/lA cultures were then incubated in a shaking









incubator (Queue Systems, Asheville, NC) at 30 rotations per minute at 370C for 24

hours. The Erwinia culture was incubated on a desktop shaking incubator (Thermolyne

RotoMix Type 50800, Dubuque, IA) at 100 rotations per minute at 25-270C for 24 hours.

The cultures were successively transferred for two days in 10 mL of fresh TSB (rif+) to

obtain uniform cell type (Beuchat et al. 2001) and incubated again at its respective

temperature for 24 hours. Three 100 mL flasks of fresh TSB (rif+) were prepared. A

loop of each serovar was used to inoculate each of the flasks. Every hour a cuvette

reading of the culture (1 mL) was obtained and read in a spectrophotometer (Shimadzu

Scientific Instruments, Inc., Model UV-1201, Japan) set at 600 X. Following three

successive hourly cuvette readings, the growth curve was concluded due to the cells

reaching the stationary phase. Additionally, every hour 1 mL of the inoculated TSB

(rif+) was serially diluted (1:10) in 9 mL tubes of sterile 0.1% peptone water (BactoTM

Peptone, Sparks, MD). The appropriate dilutions were plated out using the pour plate

technique for the Salmonella and .\/nge//a serovars and the spread plate technique for the

Erwinia carotovora using NA (rif+). Plates for the cultures were allowed to set and

inverted to be statically incubated at their respective temperatures for 24-48 hours.

Colony forming units (CFU) were observed and counted.

Preparation of Inoculum

Three days prior to the beginning of each experiment, the pathogens were revived

off of TSA (rif+) slants. Overnight transfers were performed using 10 mL tubes of TSB

(rif+) each day when growth was visible. On the day of the experiment, cells were

washed twice via centrifugation (4,000 x g, 10 minutes at 40C) using 0.1% peptone

water. Equivalent aliquots of the five serovars of Salmonella were combined as a

cocktail. The same procedure was used to make the cocktail for the two species of









,/n gellt The Erwinia carotovora, .\l/gell/ and Salmonella cultures had reached

populations of approximately 1.0 x 108 CFU/mL. When making the combined cocktails

containing either Salmonella or ,l/nge/lat with Erwinia, half of the cocktail contained the

plant pathogen while the other half consisted of the previously made cocktail of the

Salmonella or .\l/ge/llt The inoculum was serially diluted using 9 mL tubes of 0.1%

peptone water to confirm cell concentrations. The dilutions were plated in triplicate

using the pour plate technique for Salmonella and ,s/nge//Al and the spread plate technique

for Erwinia carotovora using NA (rif+).

Inoculation of Tomatoes

Intact Tomatoes

Unwashed, unwaxed mature green tomatoes, variety Florida 47, were supplied by

Six L's Packing, Inc. (Immokalee, FL) and DiMare (Tampa, FL) for all experimental

studies. Tomatoes were classified as 6x7 (formerly medium) by the Florida Tomato

Committee (Florida Tomato Committee 2005). When preparing for inoculation,

tomatoes were placed aseptically on sterile fiberglass trays with the stem scars facing

down. The tomatoes were sprayed with reagent alcohol and wiped with a Kimwipe

(Kimberly-Clark, Neenah, WI) to remove any particulate matter from the surface of the

tomato. Once clean and dry, the tomatoes were inoculated around the blossom scar area,

not directly on the blossom scar, using a Repeater Plus pipette (Eppendorf AG,

Germany). The tomato inoculation consisted often 10 ptl spots of the pathogen

suspension yielding a total inoculation of 100 ptl per whole tomato. Immediately after

inoculation, the intact tomato samples were allowed to dry completely at room

temperature on the lab bench. Once dry, the samples needed for that day were taken and

the rest of the tomatoes were placed in a Caron 6030 (Caron, Marietta, OH)









environmental humidity chamber. The Caron humidity chamber was equipped with a

Caron CRS 101 water supply system which utilized distilled water to humidify the

chamber. The temperature and relative humidity inside the chamber was continuously

controlled and displayed by a Whatlow Series 96 temperature and relative humidity

controller (Whatlow, Winona, MN).

Shave-Wounded Tomatoes

Unwashed, unwaxed mature green tomatoes were prepared for inoculation in a

similar manner as mentioned above. A sharpened knife was sterilized by spraying with

reagent alcohol and putting it through a flame. Once the knife was cool to the touch,

three shave wounds were made by slicing a 1-2 mm section around the blossom end of

the tomato. Between each shaving, the knife was sterilized and a total of three shave

wounds were made on each tomato. The bacteria were inoculated directly onto the shave

wounds using ten 10 gl spots yielding a total of 100 gl of suspension. The samples

needed for that day were taken immediately and the rest were placed in a Caron 6030

environmental humidity chamber with the same functions as mentioned previously.

Pathogen Recovery off Tomato Surfaces

Intact Tomatoes

Tomatoes were sampled from the environmental humidity chamber on Days 0, 1, 2

and 3 and recovery studies were performed. Each day's sample consisted of three

replicates. On Day 0, once the inoculum was dry and before being put in the

environmental humidity chamber, the samples for that day were immediately placed in

sterile Stomacher (Seward, West Sussex, UK) bags containing 100 mL of sterile 0.1%

peptone water. For sampling on Days 1, 2 and 3, the tomatoes were aseptically removed

from the environmental humidity chamber and individually placed into sterile









Stomacher bags containing 100 mL of sterile 0.1% peptone water. Tomato samples

were rubbed and shaken for one minute (Burnett and Beuchat 2001; Harris et al. 2001;

Zhuang et al. 1995) with rubbing action concentrated around the inoculated blossom scar

area of the tomatoes to recover any attached bacteria. The sample from each Stomacher

bag was then serially (1:10) diluted using sterile 0.1% peptone water dilution tubes. The

serial dilutions were pour-plated for the Salmonella and ./nge//At cocktails and spread

plated for Erwinia, Salmonella-Erwinia and .\ligeNll-Erwinia combinations using NA

(rif+). A negative control for the NA (rif+) was poured in duplicate to ensure the media

was not contaminated. The plates were inverted and statically incubated at their

respective temperatures for 48 hours.

Shave-Wounded Tomatoes

Tomato shave wounds were sampled on Days 0, 1, 2 and 3 and recovery studies

were performed. Each day's sample consisted of three replicates. On Day 0,

immediately after inoculating the shave wound, the knife was flame sterilized with

reagent alcohol and used to remove a slice 1-2 mm in thickness at the site of inoculation.

Once these initial samples were taken, the shave wound tomatoes were placed in the

environmental humidity chamber until future sampling. Each day, samples were cut from

the shave wounded tomato and immediately put in sterile Stomacher bags containing

100 mL of sterile 0.1% peptone water. Tomato samples were stomached (AES

Laboratoire, Comourg, France) for one minute. The samples from each Stomacher bag

were then serially (1:10) diluted using sterile 0.1% peptone water dilution tubes. The

appropriate serial dilutions for the Salmonella and ./nge//t cocktail were pour plated,

while the Erwinia, Salmonella-Erwinia and .\lgel/h,-Erwinia combinations were spread









plated using NA (rif+). A negative control of NA (rif+) was plated in duplicate to ensure

the media was not contaminated.

Uninoculated whole tomatoes and shave wounded tomatoes were used as control

samples throughout the length of the studies. The intact tomato control samples were

placed in sterile Stomacher bags of 0.1% peptone water, rubbed and shaken and serially

(1:10) diluted as mentioned previously. The shave wounded control tomatoes were

sampled using a flame sterilized knife and stomached for one minute in 0.1% peptone

water. The samples were then serially (1:10) diluted and inverted to be statically

incubated at room temperature (25-270C) for 48 hours.

Shave-Wounded Tomatoes with Combined Pathogens

Shave wounded tomatoes were inoculated with the pathogen combinations of

Salmonella-Erwinia or .\/,/ge/lt-Erwinia. Samples were obtained immediately after

inoculation on Day 0, and then approximately 24 hours later on each Days 1, 2, and 3.

All tomato samples were done in triplicate to ensure repeatability. Samples were

obtained in the same manner mentioned in the previous section. Once stomached,

samples from the tomato slices were serially diluted appropriately and plated on both NA

(rif+) and Salmonella-./irlge/ll (SS) Media. It was previously determined that Erwinia

was not capable of growth on SS Media. In determining the viable bacterial counts, the

NA (rif+) plates represented the total amount of bacteria recovered (i.e., Salmonella or

\/1Nge//At and Erwinia). The SS Media represented either Salmonella, detectable by black

colonies, or .,/Nige/li, apparent as red/orange colonies. To obtain the counts for Erwinia,

the NA (rif+) plate counts (total population) were subtracted from the SS Media plate

counts (Salmonella or .,hl/gel/i,).









Statistical Analysis

All results from the longevity studies were average logo counts logoo CFU/mL) of

recovered Salmonella, .\ligel// and Erwinia. Statistical analyses were performed with

the Statistical Analysis System (SAS Institute, Cary, NC). The GLM (general linear

model) procedure in SAS was used to evaluate the significance between environmental

conditions, time (days), the condition of the tomato surface and combinations thereof for

each pathogen utilized. Multiple comparisons were performed using the LS (least

squares) Mean method. Results with P<0.05 were considered significant in these

longevity studies. In determining significance of the shave-wounded bacteria

combination tomatoes, two tailed t-tests with P<0.05 was considered significant.














CHAPTER 4
RESULTS

Recovery studies were performed to determine the survival of Salmonella, .\/lg//lt

and Erwinia on intact and shave-wounded tomato surfaces. Inoculated fruit were

subjected to three different temperature/relative humidity conditions for three days. The

three environments consisted of standard ripening room conditions (90% RH, 200C),

Florida fall/winter production line conditions (60% RH, 27C), and optimum conditions

for the survival and growth ofErwinia carotovora subspecies carotovora (90% RH,

270C). Inoculated fruit were sampled daily to determine the survival/growth of the

bacteria. The recovery rate of these bacteria was used to assess the optimum

environmental conditions and tomato surface condition that would allow for survival and

growth. Additionally, the combination recovery studies were used to determine if there

was a correlation between the presence of Erwinia and growth or survival of the

foodborne bacteria Salmonella or .hlge/ll, over time.

Each inoculation group was sampled in groups of three to five at specific intervals;

Day 0, 1, 2, and 3. Intact tomato samples consisted of the whole tomato, while the shave-

wounded tomato samples contained a 1-2 mm shaved off portion that had been

inoculated. The recovered bacteria from each of the three replicates were counted and

averaged. Next, the data was compiled into graphs depicting the relationship between

logo CFU/mL of the bacterial survivors and time (days) for either intact or shave-

wounded tomatoes in each simulated environment.










Growth Levels of Salmonella, Shigella and Erwinia carotovora

Growth studies were performed in triplicate, for each of the five rifampicin-

resistant Salmonella serovars (Figure 4-1), two rifampicin-resistant .lnlgei//a species

(Figure 4-2) and rifampicin-resistant Erwinia carotovora (Figure 4-3).


0 1 2 3 4 5 6 7 8 9 10
Time (hours)

Figure 4-1. Average logo CFU/mL growth curves of Salmonella serovars over a 10-hour
incubation.


4 S. sonnei -
2 3 S. flexneri
2 2
1
0
0 1 2 3 4 5 6 7 8 9 10 11 12
Time (hours)

Figure 4-2. Average logo CFU/mL growth curves of.\/nge//a species over a 12-hour
incubation.











10




6
O 5
04
3 --E. carotovora
e2

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 192021
Time (hours)

Figure 4-3. Average logo CFU/mL growth curves of Erwinia carotovora over a 21-hour
incubation.

Results from these preliminary studies ensured that serovar growth rates were

equivalent to one another and a consistent inoculum suspension could be made (Figure

4-4). The cell concentration of each bacterial cocktail was estimated before beginning

any experiments. To determine the suspension concentration the cocktail was pour plated

for Salmonella and .\/ge//lt and spread plated for Erwinia using the appropriate dilutions

in triplicate on NA (rif+).


10
9

6

4


0







Figure 4-4. Maximum average logo CFU/mL counts of Salmonella serovars, .\/nge//t
species and Erwinia (rif+) after specific incubation times.









Preliminary Recovery Studies: Intact Tomato Surfaces

Before beginning the longevity studies on intact and shave-wounded tomatoes,

tomatoes were inoculated with the five-strain Salmonella cocktail, two-strain .\nge//at

cocktail or Erwinia carotovora. The intact tomatoes were sampled once the inoculum

had dried on the surface. Each tomato underwent the shake-rub method and was plated

using the pour plate method for the Salmonella and .\/nge//t and the spread plate method

for Erwinia. All bacteria were recovered from the tomato surface with no more than a

1.5 logo CFU/mL reduction from the initial inoculation concentration (Table 4-1). Intact

tomatoes that were not inoculated served as the control and were used to ensure that the

NA (rif+) media was sufficient in eliminating background bacteria.

Table 4-1. The logo CFU/mL reduction of individual bacteria off intact tomato surfaces
during preliminary studies
Log CFU/mL SD Salmonella ,ilge/lt Erwinia
Initial inoculation 5.57 + 0.02 5.08 + 0.11 4.97 + 0.17
Recovery post drying 4.47 + 0.15 3.65 + 0.35 4.94 + 0.17
Total reduction 1.10 + 0.15 1.43 0.35 0.03 0.17

Recovery of Bacteria from Tomato Surfaces

Mature green tomatoes (Florida 47) were inoculated with the individual bacterial

suspensions or combinations on both intact and shave-wounded tomatoes. Samples were

stored for three days in an environmental humidity chamber to simulate specific

parameters. Tomatoes not inoculated served as the controls and were sampled each day

to ensure the rifampicin was efficient in eliminating background microflora off the

surface or shave-wound of the tomato samples. All control tomatoes were found to be

negative.









Recovery of Bacteria from Intact Tomato Surfaces

Salmonella Recovery off Intact Tomato Surfaces

Intact tomatoes were subjected to the three simulated environments. The initial

Salmonella inoculum concentration applied to all intact tomato surfaces was

approximately 8.09 logo CFU/mL. All intact tomatoes inoculated with the Salmonella

cocktail subjected to the three simulated environments showed an overall decrease in

logo CFU/mL values from Day 0 to 3 (Figure 4-5). The smallest logo CFU/mL

reduction was shown to occur with the 90% RH, 270C conditions with a total decrease of

0.64 logo CFU/mL. The reduction of Salmonella from tomatoes subjected to the 90%

RH, 200C conditions was 4.43 logo CFU/mL, while the reduction of Salmonella off of

the tomatoes subjected to the 60% RH, 270C conditions was 3.59 logo CFU/mL.

From LS Means analysis, it was shown that the 60% RH, 270C had significantly

lower recovery than both environments kept at 90% RH (P<0.05). There were no

significant differences observed between the intact tomatoes inoculated with Salmonella

that were subjected to 90% humidity regardless of temperature (200 or 270) (P>0.05).

For all intact Salmonella studies, there were no significant differences observed between

environmental conditions on Day 0. The recovered Salmonella was significantly less

under 60% RH, 270C environmental conditions than the 90% RH, 270C and 90% RH,

200C on Day 1. All three environmental conditions were significantly different from

each other on Days 2 and 3 (P<0.05). It can be observed that Salmonella is capable of

surviving on intact tomato surfaces over three days with higher humidity and temperature

(90% RH, 270C) allowing for the most recovery.











6

S5





2 --60% RH, 27C
1 --90% RH, 27C
90% RH, 20C
0
0 1 Time (days) 2 3

Figure 4-5. Salmonella recovery logoo CFU/mL) from intact tomato surfaces in
fall/winter (60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C)
and ripening room parameters (90% RH, 200C) over three days.

Shigella Recovery from Intact Tomato Surfaces

Intact tomatoes were subjected to the same three environmental conditions as

explained above. The initial inoculation concentration applied to the intact tomatoes was

approximately 8.04 logo CFU/mL. All intact, inoculated tomatoes displayed an average

logo CFU/mL decrease over the three day study (Figure 4-6). The greatest logo

CFU/mL decrease occurred under the standard ripening room conditions (90% RH,

200C) with a 4.39 logo CFU/mL reduction. The recovery under fall/winter conditions

(60% RH, 270C) resulted in a 3.69 logo CFU/mL reduction, while the optimum

conditions for Erwinia (90% RH, 270C) resulted in a 1.36 logo CFU/mL reduction.

There was a significant increase in recovered .\lige/ll/ on Day 1 held at the

optimum conditions for Erwinia (90% RH, 270C). There was also a slight increase in

logo CFU/mL shown on Day 2 of the fall/winter conditions (60% RH, 270C). LS Means

analysis showed that there were no significant differences between the fall/winter season

conditions (60% RH, 270C) and the standard ripening room conditions (90% RH, 200C).









The recovery of .\l/nge//a under optimum conditions for Erwinia (90% RH, 270C) was

significantly greater than the other two treatments. There were no significant differences

in recovered .\l/nge// found between Days 1, 2 and 3 under fall/winter and standard

ripening room conditions. The least recovery occurred under fall/winter season

conditions (60% RH, 200C) and the standard ripening room conditions (90% RH, 200C).

.\1/n'/l t was most recoverable on intact tomato surfaces for three days under the

optimum conditions for the growth of Erwinia (90% RH, 270C).


6 60% RH, 27C
-p 90% RH, 27C
i5- 90% RH, 20'C
4



02

1

0
0 0.5 1 1.5 2 2.5 3
Time (days)
Figure 4-6. .\//ige// recovery logoo CFU/mL) from intact tomato surfaces in fall/winter
(60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C) and
ripening room parameters (90% RH, 200C) over three days.

Erwinia Recovery from Intact Tomato Surfaces

Intact tomatoes inoculated with an initial concentration of approximately 8.15 logo

CFU/mL of Erwinia carotovora were subjected to the same three environmental

conditions as explained previously. Erwinia subjected to all three environmental

conditions showed an average logo CFU/mL decrease over the three day study (Figure 4-

7). The greatest decrease occurred under standard ripening room conditions (90% RH,

200C) with a reduction of 5.71 logo CFU/mL. The recovery under fall/winter conditions








(60% RH, 270C) resulted in a 4.58 logo CFU/mL reduction, while the optimum
conditions for Erwinia resulted in a 1.36 logo CFU/mL decrease. There was a slight
increase in recovery observed on Day 2 under optimum conditions for Erwinia (90% RH,
270C).
There were no significant differences observed between the standard ripening room
conditions (90% RH, 200C) and the optimum conditions for Erwinia carotovora (90%
RH, 270C) on Day 0 according to LS Means analysis. There were no significant
differences observed between Day 1 and Day 2 for the intact tomatoes inoculated with
Erwinia. Erwinia was able to survive for the span of the three day study under its
optimum conditions for growth (90% RH, 270C) on intact tomato surfaces. The
fall/winter season conditions (60% RH, 270C) had no recovery of Erwinia after Day 0.
The lower temperature (200C) and relative humidity (60%) was an important factor in the
recoverability of the Erwinia off of intact tomato surfaces.

--60% RH, 27C
7 -- 90% RH, 27C
6 __ 90% RH, 20C
," !fri


3

1
0


0 1 2 3
Time (days)
Figure 4-7. Erwinia carotovora recovery logoo CFU/mL) from intact tomato surfaces in
fall/winter (60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C)
and ripening room parameters (90% RH, 200C) over three days.


I~ --.:Ii









Recovery of Bacteria from Shave-Wounded Tomato Surfaces

Salmonella Recovery from Shave-Wounded Tomato Surfaces

Shave-wounded tomatoes were subjected to three environmental conditions

consisting of fall/winter tomato season (60% RH, 270C), optimum conditions for Erwinia

(90% RH, 270C), and standard ripening room conditions (90% RH, 200C). The inoculum

was serially diluted to allow for growth on the wounded tomato surface. Each shave-

wound was inoculated with approximately 5.46 logo CFU/mL of the five-strain

Salmonella cocktail. All shave-wounded tomato surfaces showed an average logo

CFU/mL increase during the three day experiments (Figure 4-8). The greatest logo

CFU/mL increase occurred under standard ripening room conditions (90% RH, 200C)

with an increase of 4.00 logo CFU/mL. The optimum conditions for Erwinia (90% RH,

270C) resulted in a 3.35 logo CFU/mL increase, while the fall/winter conditions (60%

RH, 270C) showed a 1.26 logo CFU/mL increase.


7 -- 60% RH, 27C
S-_--90% RH, 27C -_



( 4



S----
0


0 1 Time (days) 2 3

Figure 4-8. Salmonella recovery logoo CFU/mL) from shave-wounded tomato surfaces
in fall/winter (60% RH, 270C), optimum conditions for Erwinia (90% RH,
270C) and ripening room parameters (90% RH, 200C) over three days.

Results from LS Means analysis showed that significantly fewer Salmonella were

recovered at 60% RH, 270C when compared to both 90% RH, 270C and 90% RH, 20C









(P<0.05). Under higher humidity conditions (90% RH) there were no significant

differences observed on Day 2 and 3, showing that humidity and not temperature is an

important factor in the survival and growth of Salmonella on wounded tomato surfaces.

Shigella Recovery from Shave-Wounded Tomato Surfaces

The initial .\l/ge/lA inoculum was serially diluted to allow for growth on the

wounded tomato surface. Shave-wounded tomatoes were inoculated with an initial

concentration of approximately 5.04 logo CFU/mL and subjected to the same three

environmental conditions as explained previously. There was no recovery observed from

any of the samples obtained under fall/winter conditions (60% RH, 27C), while the

standard ripening room (90% RH, 200C) and optimum conditions for Erwinia (90% RH,

270C) resulted in a slight logo CFU/mL increase (Figure 4-9). .\/ngel/l subjected to

standard ripening room conditions resulted in a 2.35 logo CFU/mL increase and the

optimum conditions for Erwinia resulted in a 1.28 logo CFU/mL increase over the span

of three days. On Day 1 of the Erwinia optimized conditions (90% RH, 27C), there was

a relatively large increase in logo CFU/mL for .\lge//A before decreasing slightly on

Day 2 and remaining constant for the remainder of the study.

There were no significant differences observed between the standard ripening room

conditions (90% RH, 200C) and the optimum conditions for Erwinia (90% RH, 270C)

over three days. The recovery of.\,/ge//At under fall/winter conditions was significantly

less than the environmental conditions held at higher humidity (90%). Under the

fall/winter season conditions, a possible explanation for the lack of recovered cells may

be that the inoculum did not sufficiently attach to the wounded tomato surface before

being placed in the environmental humidity chamber. Once inside the humidity chamber,






40


the wounded tomato surfaces showed visible desiccation within one day preventing the

survival of any .\/nge//t that may have been present.


5

04 -

3


-*-60% RH, 27C
r 1 ----90% RH, 27C
.-A 90% RH, 20C
0 0-
0 1 Time (days) 2 3

Figure 4-9. .liigell/ recovery logoo CFU/mL) from shave-wounded tomato surfaces in
fall/winter (60% RH, 270C), optimum conditions for Erwinia (90% RH, 270C)
and ripening room parameters (90% RH, 200C) over three days.

Erwinia Recovery from Shave-Wounded Tomato Surfaces

Shave-wounded tomatoes were kept in three separate environmental conditions for

three days as explained previously. The initial concentration of Erwinia carotovora was

serially diluted and applied to all shave-wounded tomatoes at a concentration of

approximately 5.64 logo CFU/mL. The largest increase of Erwinia carotovora

recovered occurred under optimum conditions (90% RH, 270C) resulting in a 6.59 logo

CFU/mL increase (Figure 4-10). The fall/winter conditions resulted in a 2.59 logo

CFU/mL increase, while the standard ripening room conditions resulted in a 5.93 logo

CFU/mL increase.

There were no significant differences observed between the standard ripening room

conditions (90% RH, 200C) and the optimum conditions for Erwinia (90% RH, 270C)

according to LS Means analysis. The recovery of Erwinia under fall/winter conditions

(60% RH, 270C) was significantly less than the environments held at higher humidity's









(90% RH, 270C and 90% RH, 200C). The lower humidity (60%) used for fall/winter

conditions allowed for some growth of Erwinia but did not show the exponential growth

displayed by those environments kept at a higher humidity (90%). A high relative

humidity is important in allowing for survival and growth ofErwinia carotovora.


10 --,60% RH, 27C
S9 .- 90% RH, 27C
-- 90% RH, 200C






0
5
-


1 ^
0 <^ --------------.

0 1 Time (days) 2 3
Figure 4-10. Erwinia carotovora recovery logoo CFU/mL) from shave-wounded tomato
surfaces in fall/winter (60% RH, 270C), optimum conditions for Erwinia
(90% RH, 270C) and ripening room parameters (90% RH, 200C) over three
days.

Recovery of Combined Bacteria from Shave-Wounded Tomato Surfaces

Recovery of Salmonella and Erwinia in Fall/Winter Season Conditions

Shave-wounded tomatoes were kept in the environmental humidity chamber for the

length of the study at the fall/winter season conditions (60% RH, 270C). The initial

concentration of the Salmonella-Erwinia combination was serially diluted and applied to

the shave-wounds at a concentration of approximately 5.35 logo CFU/mL. The

concentration of the inoculation was reduced to allow for growth. Both the Salmonella

and Erwinia counts increased throughout the three-day study resulting in a 3.99 logo

CFU/mL increase of Erwinia and a 2.72 logo CFU/mL increase of Salmonella (Table 4-

2).









According to the two tailed t-test, there were significant increases observed for

Salmonella between Day 1 and 2 of the study. The combined bacteria under fall/winter

season conditions allowed for a greater amount of growth of both microorganisms than

when individually inoculated onto the wounded tomato surface as seen at the Day 2

sampling period and beyond. These effects show that the bacteria may be creating more

favorable environments for each other by Erwinia increasing the moisture content,

releasing nutrients and Salmonella possibly providing a protective covering to the lower

humidity (60%) when combined.

Table 4-2. Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in
fall/winter (60% RH, 270C) conditions over three days.
Loglo CFU/mL Salmonella Salmonella Erwinia Erwinia
+ SD individually recovered in individually recovered in
combination combination
with Erwinia with
Salmonella
Day 0 0.70 1.09a 1.85 0.52a 0.26 0.58a 2.13 0.66b
Day 1 0.92 0.95a 1.59 1.34a 1.98 1.14a 2.67 1.14a
Day 2 1.42 + 1.37a 4.22 0.97b 1.91 + 2.01a 6.00 + 0.54b
Day 3 2.00 + 1.68a 4.56 0.80b 2.85 + 2.67a 6.11 + 0.37b
Note: Letters (a,b) following Loglo CFU/mL SD among bacteria (i.e., Salmonella vs.
Salmonella combined) in rows are significantly different.

Recovery of Salmonella and Erwinia in Optimum Conditions for Erwinia

Shave-wounded tomatoes were kept in the environmental humidity chamber for the

length of the study at the conditions considered optimal for Erwinia carotovora (90%

RH, 270C). The initial concentration of the applied inoculum of Salmonella-Erwinia was

approximately 4.87 logo CFU/mL. This inoculum concentration was obtained by

serially diluting the stock culture, which would allow for sufficient growth on the

wounded tomato surface. The Salmonella and Erwinia grew substantially over the period









of three days resulting in a 4.94 logo CFU/mL increase of Erwinia and a 4.81 logo

CFU/mL increase of Salmonella (Table 4-3).

There were no significant differences observed in growth Salmonella when

comparing the individual counts and those recovered in combination with Erwinia on

Days 1,2 or 3. In contrast, the Erwinia grew to a significantly lesser amount when

combined with the Salmonella. This could be due to competition between the

microorganisms Salmonella and Erwinia.

Table 4-3. Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in
optimized conditions for Erwinia carotovora (90% RH, 270C) over three
days.
Loglo CFU/mL Salmonella Salmonella Erwinia Erwinia
+ SD individually recovered in individually recovered in
combination combination
with Erwinia with
Salmonella
Day 0 1.98 + 0.23a 1.45 0.38b 2.21 0.08a 1.65 + 0.46a
Day 1 4.27 0.47a 5.18 + 0.98a 5.92 + 0.32a 5.61 0.99a
Day 2 5.54 1.39a 5.20 1.66a 7.93 0.17a 6.34 1.10b
Day 3 5.32 + 0.23a 6.24 1.21a 8.80 +0.17a 6.59 0.45b
Note: Letters (ab) following Loglo CFU/mL SD among bacteria (i.e., Salmonella vs.
Salmonella combined) in rows are significantly different.

Recovery of Salmonella and Erwinia in Standard Ripening Room Conditions

Shave-wounded tomatoes were kept in an environmental humidity chamber for

three days set to the conditions of a standard ripening room (90% RH, 200C). The stock

culture was serially diluted before being inoculated onto the shave-wounded tomatoes to

allow for growth. The initial concentration of the Salmonella-Erwinia combination

inoculum was 5.14 logo CFU/mL. The Salmonella and Erwinia grew daily with the

biggest increase seen between Day 0 and Day 1 (Table 4-4). The Erwinia had a total

increase of 6.26 logo CFU/mL while the Salmonella had a 4.18 logo CFU/mL increase

over three days.









Growth of Salmonella combined with Erwinia was not significantly different from

Salmonella individually inoculated onto shave-wounded tomatoes except on Day 3;

growth was significantly less. The same pattern of significance was seen for Erwinia.

Both Salmonella and Erwinia were recovered in lower numbers when combined under

standard ripening room conditions (90% RH, 200C). Although growth was observed

when combined, the microorganisms may be competing with each other for nutrients and

available space resulting in less growth.

Table 4-4. Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in
standard ripening room conditions (90% RH, 200C) over three days
Loglo CFU/mL Salmonella Salmonella Erwinia Erwinia
+ SD individually recovered in individually recovered in
combination combination
with Erwinia with
Salmonella
Day 0 2.08 0.17a 0.92 0.85a 2.12 0.18a 1.89 1.13a
Day 1 3.68 + 1.03a 4.83 0.07a 6.04 0.11a 5.44 1.07a
Day 2 5.53 0.78a 4.55 0.06a 7.05 0.32a 6.35 0.41a
Day 3 6.03 0.48a 5.08 0.33b 8.04 + 0.10a 7.35 0.04b
Note: Letters (ab) following Loglo CFU/mL SD among bacteria (i.e., Salmonella vs.
Salmonella combined) in rows are significantly different.

Recovery of Shigella and Erwinia in Fall/Winter Season Conditions

Shave-wounded tomatoes were placed in an environmental humidity chamber set to

simulate Florida fall/winter tomato season conditions (60% RH, 270C) for three days.

The concentration of the lngel//,t-Erwinia combination applied to the shave-wounded

tomatoes was 5.00 logo CFU/mL. On Day 1 and Day 2 of the study, there was no

.\l/ge/ll recovered (Table 4-5). Unexpectedly on Day 3, there was a considerable

amount of ,\s/ge//t recovered from the shave-wounded tomatoes. This was most

probably due to the increase in moisture available after the Erwinia carotovora had time

to break down tomato cell walls and release fluids. There was an increase seen in the









Erwinia throughout the study resulting in a 4.27 logo CFU/mL increase, while the

\/nge//At had a total increase of 3.92 logo CFU/mL.

There were no significant differences between Day 1 and Day 2 for .\ igel//A since

none was recoverable on standard plating methods. Over three days, both ./nge//lt and

Erwinia displayed significant increases (P<0.05). Erwinia displayed greater growth

when inoculated in combination with ./nge//lt on the wounded tomato surface. .\/nge//t

was able to survive in combination with Erwinia under the lower relative humidity

conditions unlike when inoculated alone on the wounded tomato surfaces. The Erwinia

was able to provide greater moisture availability by breaking down the plant tissues of the

tomato and releasing fluids, allowing the ./nge//At to survive and then grow when

conditions were most favorable. The .\lngel//t may have provided a protective barrier for

the Erwinia under the low humidity conditions (60%). Under these conditions, there is a

synergistic effect observed between .\tige//At and Erwinia in creating an optimum

environment for themselves to allow for survival and growth.

Table 4-5. Recovered \l/igell// and Erwinia from shave-wounded tomato surfaces in
fall/winter season conditions (60% RH, 270C) over three days. ND: Non-
detectable by lowest dilution plating methods.
Loglo CFU/mL lge/t lget Erwinia Erwinia
+ SD individually recovered in individually recovered in
combination combination
with Erwinia with

Day 0 NDa 1.19 0.22b 0.26 0.58a 1.93 0.22b
Day 1 NDa NDB 1.98 + 1.14a 1.96 0.74a
Day 2 NDa ND 1.91 + 2.01a 2.53 2.04"
Day 3 NDa 5.08 0.75b 2.85 + 2.67a 6.20 0.22b
Note: Letters (ab) following Logio CFU/mL + SD among bacteria (i.e., \/nge//lt vs.
\/nge"//At combined) in rows are significantly different.









Recovery of Shigella and Erwinia under Optimum Conditions for Erwinia

Shave-wounded tomatoes were placed in an environmental humidity chamber to

simulate the optimized conditions for growth ofErwinia carotovora (90% RH, 270C).

The stock culture was serially diluted before inoculating the shave-wounded tomatoes to

allow for growth. The .\ligell/,-Erwinia combination inoculum had an initial

concentration of 4.87 logo CFU/mL. There was a steady increase in growth seen

throughout the study for both the .\/nge/hi and Erwinia (Table 4-6). By Day 3, the

./nlge/t had an increase of 4.84 logo CFU/mL while the Erwinia had an increase of 6.55

logo CFU/mL.

Table 4-6. Recovered .l/nge/,t and Erwinia from shave-wounded tomato surfaces under
optimum conditions for Erwinia carotovora (90% RH, 270C) over three days.
Loglo CFU/mL ,/ng t Erwinia Erwinia
+ SD individually recovered in individually recovered in
combination combination
with Erwinia with

Day 0 1.69 + 0.29a 0.69 + 0.58b 2.21 0.08a 1.00 + 0.89
Day 1 3.49 0.73a 0.92 + 0.27b 5.92 + 0.32a 4.02 + 0.58b
Day 2 2.99 1.32a 3.22 1.63a 7.93 0.17 6.19+ 1.06b
Day 3 2.97 + 1.49a 5.51 0.39b 8.80 0.17a 7.55 0.55b
Note: Letters (ab) following Loglo CFU/mL SD among bacteria (i.e., .\/nge/lt vs.
.\/ngl/lt combined) in rows are significantly different.

.\l1ge//At exhibited significant growth (P<0.05) over time in combination with

Erwinia. Growth of Erwinia in combination with ./nhge/lt was significantly less in

combination with .\l/gell// Under these environmental conditions, the .\l/gel,// grew to a

much higher population when inoculated onto the wounded tomato surface in

combination with Erwinia. There was significantly less growth of Erwinia observed on

Day 3 when combined with .\/nge//l These results show that the Erwinia is allowing for

more favorable conditions for the .\l/ge//At survival and growth by increasing moisture









availability and possibly releasing nutrients previously not available before tissue

maceration.

Recovery of Shigella and Erwinia in Standard Ripening Room Conditions

Shave-wounded tomatoes were subjected to ripening room conditions (90% RH,

200C) for a span of three days and inoculated with a .,/nige/l-Erwinia combination

inoculum with an initial concentration of 4.88 logo CFU/mL. There were significant

increases (P<0.05) for both Erwinia and ,/nge/l/h over the span of the three day study

(Table 4-7). The total increase of \/nge//At was 3.35 logo CFU/mL while the Erwinia

increased 6.39 logo CFU/mL.

Under standard ripening room conditions, l/nge/lit and Erwinia recovered in

combination was significantly less than when recovered individually. Since

environmental conditions were not considerably stressful for either organism, it may be

possible that the microorganisms were recovered at lesser concentrations when inoculated

together due to competition with limited surface area for growth.

Table 4-7. Recovered .\ligel // and Erwinia from shave-wounded tomato surfaces under
standard ripening room conditions (90% RH, 200C) over three days.
Loglo CFU/mL /nget /et Erwinia Erwinia
+ SD individually recovered in individually recovered in
combination combination
with Erwinia with

Day 0 1.63 + 0.33a 0.35 0.58 2.12 + 0.18a 1.82 1.00a
Day 1 2.62 0.03a 2.64 0.19a 6.04 + 0.11a 2.91 0.12b
Day 2 3.10 + 0.17a 2.94 0.13a 7.05 0.32a 6.52 + 0.48a
Day 3 3.98 + 0.70a 3.68 0.21a 8.04 0.10a 7.52 0.18b
Note: Letters (,b) following Loglo CFU/mL SD among bacteria (i.e., .\/nge/lt vs.
./n"gel'/ combined) in rows are significantly different.














CHAPTER 5
DISCUSSION

Consumption of fresh fruits and vegetables has greatly increased over the past 10

years (Lucier et al. 2001). The produce industry is continually confronted with the threat

of microbial food safety hazards. Many steps are taken during the harvesting, processing

and distribution of fresh produce, with the possibility of pathogenic contamination

increasing significantly along the way. Environmental factors such as temperature and

relative humidity have a large impact on the quality of fruits and vegetables along with

the survival capacity of potential bacteria (Allen 2003). In recent years, multiple

foodborne illnesses of Salmonella and .\l/ge/l/ have been associated with the

consumption of contaminated produce, specifically tomatoes (Beuchat 1995; Brackett

1999).

This study utilized a five-serovar rifampicin-resistant Salmonella cocktail, a two-

species rifampicin-resistant .\l/geii/l cocktail and a rifampicin-resistant strain ofErwinia

carotovora. These cocktails were applied individually and in combination to intact and

shave wounded tomato surfaces. The tomatoes were subjected to different temperature

and relative humidity combinations that simulated the conditions found in a standard

ripening room (90% RH, 200C), fall/winter production conditions (60% RH, 270C) and

optimum conditions for the growth of Erwinia (90% RH, 270C). Recovery of the

bacteria from the intact tomatoes was performed by placing them into 100 mL of 0.1%

peptone water and applying a rub-shake method (Burnett and Beuchat 2001), while the









bacteria recovered from the shave wounded tomatoes were obtained by placing the sliced

off wounded sections into 100 mL of 0.1% peptone water and stomaching for one minute.

It has been recommended that a Salmonella cocktail should contain a minimum of

five strains at approximately equal populations (CFSAN-FDA 2001). The five

Salmonella enteritidis serovars selected for this study were S. Agona, S. Gaminara, S.

Michigan, S. Montevideo and S. Poona. The serovars S. Agona, Gaminara and Michigan

were isolated from fresh produce or their products (orange juice). S. Montevideo and

Poona serovars were human isolates associated with fresh produce outbreaks. The two

.\l1gell// species selected for this study were S. flexneri and S. sonnei since they are the

species most responsible for foodborne outbreaks. Erwinia carotovora subspecies

carotovora was selected for this study because it is know to cause soft rot in many types

of produce. All microorganisms were adapted to 200 [g/mL of rifampicin. Rifampicin

was selected because it is a stable marker and is effective in isolating bacteria from

inoculated fruits that have natural background microflora.

Growth characteristics for all microorganisms were evaluated by conducting

growth studies. For the Salmonella serovars, S. Agona was observed to have the highest

population at the end the 10-hour incubation period (37C), but there was less than a 0.6

logo CFU/mL difference between the population of S. Agona and the serovar with the

lowest growth level, S. Gaminara (Figure 4-1). All five serovars achieved counts of at

least 1.0 x 108 CFU/mL after 10 hours of incubation. For the .\/ge/lal species, S. flexneri

grew to the greatest population at the end of the 12-hour incubation period (37C), with

less than a 0.5 logo CFU/mL difference between the S. flexneri and S. sonnei growth

(Figure 4-2). Both species achieved counts of at least 1.0 x 108 CFU/mL after the 12-









hour incubation. Erwinia carotovora grew to a population of at least 1.0 x 108 CFU/mL

after the 21-hour incubation (Figure 4-3). These results showed that an acceptable

inoculum could be prepared from each of the bacteria. Prior to each study, appropriate

serial dilutions of the inoculum were pour-plated to determine the viable population of

Salmonella and .\/ge//ll while the Erwinia was spread plated. These counts for each

prepared inoculum for all experiments conducted also showed little variation between

one another. All bacterial suspensions were determined to contain viable populations at

1.0 x 108 CFU/mL (Figure 4-4).

Recovery of Bacteria from Intact Tomato Surfaces

Intact tomatoes were subjected to three simulated environments including ripening

room conditions (90% RH, 200C), fall/winter tomato production season parameters (60%

RH, 270C) and optimum conditions for the growth ofErwinia carotovora (90% RH,

27C). Intact tomatoes have a firm surface that can withstand moderate rubbing and

agitation so bacterial recovery from the tomato surface was performed by using a rub-

shake method. This rub-shake method was chosen because it is the most effective

technique for removing microorganisms from the surfaces of whole fruits and vegetables

(FDA 2001). Intact, whole, unblemished tomatoes were specifically chosen for these

inoculation studies. Spot inoculation of the cocktails was utilized because it allows for a

known number of cells to be administered to the produce.

Results from this study were similar to the findings of Guo et al. (2001) in that

Salmonella populations decreased over time on intact tomato surfaces in all simulated

environmental conditions (Figure 4-5). Salmonella was least recovered under standard

ripening room conditions (90% RH, 200C), while recovery under fall/winter production

conditions (60% RH, 270C) was not significantly greater (Figure 4-5). Between Day 0









and Day 1, the largest logo CFU/mL decrease of 2.53 logo CFU/mL was seen under

fall/winter production conditions (60% RH, 27C). Whereas, the most recovered

Salmonella from intact tomato surfaces was seen under optimum conditions for Erwinia

carotovora with only a 0.64 logo CFU/mL decrease over three days. The higher

humidity and temperature settings for the optimum conditions for growth of Erwinia

allowed for the greatest recovery of the Salmonella cocktail. Lower humidity, such as

what was used for the fall/winter production season conditions and the lower temperature

used for the standard ripening room conditions limited the survival of the Salmonella

cocktail on the intact tomato surface.

Similar to the intact tomato studies inoculated with the Salmonella cocktail, there

was an overall decrease of recovered .\,/nge/l, over time (Figure 4-6). Both the standard

ripening room conditions (90% RH, 200C) and the fall/winter production conditions

(60% RH, 270C) produced significant decreases in recovered cells from Day 0 to Day 1.

There was an unexpected increase of .\/nge//a recovered on Day 1 under optimum

conditions for the growth of Erwinia (90% RH, 27C), and also on Day 2 under

fall/winter production conditions (60% RH, 27C). These specific time points had

relatively large error bars indicating that human error in counting appropriate dilution

plates or contaminated plates from improper handling techniques may have been possible

factors in the elevated counts observed. Humidity and temperature are both important in

the survival of.\h/ge//At on the surface of intact tomatoes. As seen with the Salmonella

cocktail, the greatest recovery of cells was seen under conditions of higher humidity and

temperature, or those of optimum conditions for the growth of Erwinia (90% RH, 270C).









As with Salmonella and .l/ngel//i, Erwinia carotovora also decreased over time

(Figure 4-7) on the intact tomato surface. The optimum conditions for growth of Erwinia

(90% RH, 270C) allowed the most recovery from the intact tomato surface. The standard

ripening room conditions (90% RH, 200C) and the fall/winter production conditions

(60% RH, 270C) had the least recovered Erwinia over time. There was an unexpected

increase seen on Day 2 under optimum conditions for the growth of Erwinia (90% RH,

270C). Additionally, the standard deviation on Day 2 under standard ripening room

conditions was very large. One explanation for this event could be that one of the

samples taken that day had a break in the skin that had not been seen before spot

inoculation and had developed soft rot. This resulted in the growth of Erwinia on what

originally appeared to be an intact tomato. Humidity and temperature plays an important

role in the survival of Erwinia on intact tomato surfaces. At its optimum conditions,

Erwinia was able to survive the best, whereas the lower humidity or lower temperature

conditions significantly affected the survival of Erwinia after initial inoculation.

The bacteria Salmonella, .\ligell/, and Erwinia seem to survive but not proliferate

on intact tomatoes under high humidity and room temperature (27C). The bacteria used

in this research were observed to decline on the outer surface of tomatoes over time and

the rate of reduction seemed to be related to both temperature and humidity. Growth of

the three organisms on intact tomatoes was not seen. Foodborne bacteria are not capable

of producing the enzymes necessary to breakdown the protective outer barriers on

produce, which prevents nutrients from becoming available (Wells and Butterfield 1999).

Although Erwinia does have the capability to produce these enzymes, unless there is a

break in the tomato skin or other opening, it cannot produce soft rot (Toth et al. 2003).









Salmonella was recovered from tomato surfaces in all three simulated environments and

this indicates that the pathogen can survive on tomato surfaces and should be a concern in

the fresh produce industry. Cross-contamination is possible if contaminated fruit are

mixed in with other clean fruit, which may pose a threat to food safety.

Recovery of Bacteria from Wounded Tomato Surfaces

Shave wounded tomatoes were stored for three days in the same environmental

conditions as mentioned previously. In accordance with findings by Wells and

Butterfield (1997), the Salmonella cocktail was able to not only survive on the wounded

tomato surface, but was capable of growing. The greatest amount of growth was seen

under standard ripening room conditions (90% RH, 200C), closely followed by the

optimum conditions for the growth ofErwinia (90% RH, 270C) (Figure 4-8). The

slowest rate of growth occurred under fall/winter production season conditions (60% RH,

270C) showing that even at lower humidity, Salmonella is capable of not only surviving

on the surface of wounded tomatoes, but growing. In agreement with Beattie and Lindow

(1999), these findings suggest that a high relative humidity is imperative in the

exponential growth of Salmonella on wounded tomato surfaces.

Similar to Salmonella on shave-wounded tomatoes, the .\lige/lat cocktail was able

to survive and grow on the wounded tomato surface under higher humidity conditions

(Figure 4-9). On Day 2 under optimum conditions for the growth ofErwinia (90% RH,

27C), there was a slight decrease in recovered cells, though the standard deviations seen

were substantially large indicating other factors influencing growth and recovery. Unlike

the Salmonella cocktail, .l/ge/lit was not recoverable at any amount under the fall/winter

production season conditions (60% RH, 270C). A possible explanation for this is that the

inoculum may desiccate to a point which would not support growth and/or recovery. In









addition, at each subsequent day of the study, it was observed that the wound sites on the

tomatoes kept at the lower humidity exhibited excessive desiccation and shriveling at the

site. When inoculated by itself, .\l/ge/ll requires a relatively high humidity to survive on

wounded tomato surfaces.

As expected, Erwinia carotovora was capable of survival and growth on wounded

tomato surfaces at all environmental conditions (Figure 4-10). The greatest amount of

growth occurred under their optimum condition (90% RH, 270C) which was not

significantly greater than standard ripening room conditions (90% RH, 200C). The

slowest rate of growth occurred under conditions of lower humidity, the fall/winter

production conditions (60% RH, 27C). As seen with the wounded tomatoes for both the

Salmonella and .\/ge/All cocktails, at the lower humidity the wound sites exhibited some

desiccation preventing the exponential growth seen at the higher humidity conditions.

Although 200C is colder than what is preferential for the growth of Erwinia, under high

humidity, the conditions remained favorable enough that the plant pathogen could survive

and grow.

The bacteria Salmonella and Erwinia are capable of growing and being recovered

from wounded tomato surfaces at all three environmental conditions used. .l/nge/lt was

able to proliferate at the higher humidity conditions, but not at a humidity of 60%. If

there were a break in the surface of the tomato, there is a distinct possibility that a

foodborne pathogen could survive and cause illness under the right conditions. Tomatoes

with Salmonella and .l/nge/lat present would not show signs of being contaminated as

opposed to soft rot seen with Erwinia. Since Erwinia produces a characteristic soft rot, it









is possible for packinghouses to carefully observe the tomato fruit for soft rot and carry

out culling practices to prevent the spread of this plant pathogen among the tomato bins.

Recovery of Combined Bacteria from Wounded Tomato Surfaces

In accordance with the findings of Wells and Butterfield (1997), Salmonella grew

to higher levels in combination with Erwinia under fall/winter conditions and optimum

conditions for Erwinia (Table 4-2; Table 4-3) when compared to Salmonella grown

individually. The greatest amount of Salmonella growth was seen under optimum

conditions for the growth of Erwinia (90% RH, 270C), while the greatest amount of

growth for Erwinia was seen under standard ripening room conditions (90% RH, 200C)

(Table 4-4). It can be determined from this research that the bacterial combination

studies involving Salmonella and Erwinia did not grow at optimum levels under one

specific environmental condition. Since both environmental conditions are acceptable for

the growth of the bacteria, it may be possible that they are competing for nutrients and

the available space to multiply. The surface pH of a tomato is approximately 4.5 (Guo et

al. 2001). When Erwinia begins its soft rot pathogenesis on tomatoes the pH may

increase slightly. An increase in pH on the tomato surface would be more favorable to

Salmonella, which prefers a neutral pH, and could explain the greater amount of growth

seen of Salmonella on wounded tomato surfaces when combined with Erwinia.

A possible explanation for the lack of coordinated growth of both the Salmonella

and Erwinia in combination on wounded tomato surfaces under each environmental

condition could be that quorum sensing is taking place (Smith et al. 2003). Since the

bacteria are competing for nutrients in a limited amount of space, it is possible that either

Salmonella or Erwinia has reached a quorum and is entering survival mode until

conditions become favorable again. If these studies were to continue for several more









days, it would be possible to observe if any competition among bacteria existed, resulting

in the death of one bacteria and growth of another.

The \ligel//A and Erwinia combination studies all displayed significant growth over

three days under all environmental conditions (Tables 4-5; 4-6; 4-7). The greatest

amount of growth for both .\ligel//A and Erwinia was seen at the 90% RH, 270C growth

parameters. This was to be expected since Erwinia should thrive under those conditions,

and ,/nhgel/h is capable of growth under high humidity and temperature conditions as

seen in previous studies. When inoculated alone, \/ngel//t is not capable of survival on

wounded tomato surfaces under fall/winter production season conditions (60% RH,

270C). By contrast, when combined with the soft rot plant pathogen Erwinia, .\/nhge/t

was capable of survival as well as growth. ./ngel/lt was recovered on Day 0 under low

humidity conditions when combined with Erwinia although there were no countable

levels of /ge//At on Day 1 and 2. Day 3 showed growth of ,/n'ge//At at levels higher than

the wounded tomatoes inoculated with .l/nge/lt alone at higher humidity conditions. It is

possible that the ,hl/gel/h was still present on the surface of the shave-wounded tomato

but was at such low levels that it was below detection for standard plating techniques.

Future studies could repeat these conditions utilizing a most probable number

method (MPN) assay to determine if there are viable cells of .\/ge//At present during Day

1 and 2. If the initial pH of the wounded tomato surface, approximately 4.5, was

preventing the growth of.\/nge//i, the enzymes produced during the soft rot pathogenesis

of Erwinia can increase the pH to a more favorable level of 5.5. Although this is not a

neutral pH which .l/nge/lt favors, there is less stress placed on the ,/nge//At cells since the

pH is slightly higher.









Salmonella and .\nlge//Al may be able to grow to greater levels when Erwinia

carotovora is present because the soft rot increases the surface area that these pathogens

can colonize. Available water and damaged tomato cells resulting from tissue maceration

can greatly increase the area that becomes permeated with fluid allowing for movement

and multiplication of bacteria. Under normal circumstances, the relatively rigid tomato

cell walls and definite intercellular spaces would limit bacterial motility specifically to

areas with cell fluid present from the initial shave wound injury. It may be possible that

bacteria can move on a moist surface if the intercellular spaces were congested with fluid

resulting from soft rot.

The opportunity for foodborne illness caused by Salmonella and ./nge//lt on fresh

produce, specifically tomatoes, can be greatly enhanced in the presence of the soft rot

organism Erwinia carotovora. All three environmental conditions allowed for the

survival and growth of the bacteria on the wounded tomato surface when in combination.

To ensure food safety, it is important to remove any damaged fruit to prevent the spread

of soft rot and the possibility of foodborne illness. It would be necessary to consider that

if a tomato appears to have soft rot from Erwinia carotovora, it could possibly be

contaminated with a foodborne pathogen such as Salmonella or .\l/ge/lA, especially when

the produce is held at higher temperature and humidity. Since Salmonella and .l/ge//At

have a relatively low infectious dose, it is important to inspect fresh produce for signs of

soft rot and prevent the possibility of foodborne illness.














CHAPTER 6
CONCLUSION

All objectives of this study were completed. Growth rates of five rifampicin

resistant Salmonella enterica serovars, two .\ilge/la species and Erwinia carotovora were

established and it was determined that appropriate cocktails could be made. Salmonella,

.\iigell/ and Erwinia were successfully recovered from intact tomatoes as well as shave-

wounded tomatoes held at all three environmental conditions.

It was observed that the individual microorganisms survived best on intact

tomatoes held at optimum conditions for the growth of Erwinia (90% RH, 270C). The

individual pathogens Salmonella and .\nge/All exhibited the most growth on shave

wounded tomatoes under standard ripening room conditions (90% RH, 200C), while the

Erwinia had the greatest amount of growth under its optimum conditions (90% RH,

270C). The lower humidity seen during the fall/winter season condition (60% RH, 270C)

caused desiccation of the wounds on the shave-wounded fruit, resulting in lower recovery

of the bacteria.

In combined microorganism studies, Salmonella exhibited the greatest amount of

growth with an increase of 4.81 logo CFU/mL under optimum conditions for the growth

of Erwinia (90% RH, 27C), while the greatest growth of Erwinia occurred under

standard ripening room conditions (90% RH, 200C) with an increase of 6.26 logo

CFU/mL. Under optimum conditions for the growth of Erwinia, the logo CFU/mL

counts for Salmonella and Erwinia were relatively similar, 4.81 logo CFU/mL and 4.94

logo CFU/mL respectively, possibly showing an antagonistic relationship between the









two bacteria at those environmental conditions. The least amount of growth occurred

under fall/winter season conditions (60% RH, 270C) for both the Salmonella and

Erwinia. These results show that relative humidity plays an important role in the growth

of Salmonella and Erwinia.

In combined bacteria studies involving .\l ge//Al and Erwinia, the optimum

conditions for growth of Erwinia (90% RH, 270C) allowed for the most growth for both

bacteria resulting in a 4.84 logo CFU/mL increase of .\/ige/At and a 6.55 logo CFU/mL

increase of Erwinia. The growth levels for ./nge//lt and Erwinia were slightly less under

standard ripening room conditions (90% RH, 200C). The least growth was observed

under the fall/winter tomato season conditions (60% RH, 270C). These results show that

relative humidity, and not necessarily temperature, plays an important role in the survival

and growth of both Erwinia and .\/hge//lt on wounded tomato surfaces.















APPENDIX A
SALMONELLA STATISTICS

Salmonella 60% RH, 270C

------------------------------ day=0 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 2 combo wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


R-Square
0.361841


Sum of
Squares
3.26669181
5.76128540
9.02797721


Coeff Var
66.53039


Mean Square
3.26669181
0.72016068


Root MSE
0.848623


FValue Pr>F
4.54 0.0658


logct Mean
1.275542


DF Type III SS
1 3.26669181


Mean Square F Value Pr > F
3.26669181 4.54 0.0658


t Tests (LSD) for logct
NOTE: This test controls the Type I comparisonwise error rate, not the
experimentwise error rate.

Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 0.720161
Critical Value oft 2.30600
Least Significant Difference 1.2377
Means with the same letter are not significantly different.


t Grouping Mean
A 1.8471


N x
5 combo


A
A 0.7040 5 wounded


Source









----------------------------------- day=1 ------------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 2 combo wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
Squares
1.15543332
10.70946451
11.86489783


Mean Square
1.15543332
1.33868306


FValue Pr>F
0.86 0.3800


R-Square Coeff Var Root MSE
0.097382 92.19238 1.157015


DF Type III SS
1 1.15543332


Mean Square
1.15543332


FValue Pr>F
0.86 0.3800


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 1.338683
Critical Value oft 2.30600
Least Significant Difference 1.6874

t Grouping Mean N x
A 1.5949 5 combo
A
A 0.9151 5 wounded


------------------------------- day=2 --------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 2 combo wounded


Number of Observations Read
Number of Observations Used


Source
x


logct Mean
1.255000











Source
Model
Error
Corrected Total


Sum of
DF Squares Mean Square F Value Pr > F
1 19.56421411 19.56421411 14.06 0.0056
8 11.13356508 1.39169563
9 30.69777918


R-Square Coeff Var Root MSE logct Mean
0.637317 41.83385 1.179702 2.819969


DF Type III SS
1 19.56421411


Mean Square
19.56421411


FValue Pr>F
14.06 0.0056


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 1.391696
Critical Value oft 2.30600
Least Significant Difference 1.7205


t Grouping
A
B


Mean N x
4.2187 5 combo
1.4212 5 wounded


--------------------- --------- day=3------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 2 combo wounded

Number of Observations Read 10
Number of Observations Used 10
Dependent Variable: logct
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 1 16.41247704 16.41247704 9.94 0.0136
Error 8 13.21387226 1.65173403
Corrected Total 9 29.62634930


R-Square Coeff Var Root MSE logct Mean
0.553982 39.22735 1.285198 3.276281


DF Type III SS
1 16.41247704


Mean Square F Value Pr > F
16.41247704 9.94 0.0136


Source
x


Source









t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 1.651734
Critical Value oft 2.30600
Least Significant Difference 1.8744


t Grouping
A
B


Mean
4.5574
1.9952


N x
5 combo
5 wounded
Salmonella 90% RH, 200C


------------------------------- day=0--------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 2 combo wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
Squares
1.88614749
1.53811471
3.42426220


Mean Square
1.88614749
0.38452868


FValue Pr>F
4.91 0.0911


R-Square Coeff Var Root MSE
0.550819 41.98408 0.620104


DF Type III SS
1 1.88614749


Mean Square F Value Pr > F
1.88614749 4.91 0.0911


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.384529
Critical Value of t 2.77645
Least Significant Difference 1.4057


t Grouping Mean
A 2.0377


N x
3 wounded


A 0.9163 3 combo


Source
x


logct Mean
1.476997









----------------------------------- day= ------------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 2 combo wounded

Number of Observations Read 6
Number of Observations Used 6

Dependent Variable: logct
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 1 1.98605237 1.98605237 3.74 0.1254
Error 4 2.12624746 0.53156186
Corrected Total 5 4.11229983

R-Square Coeff Var Root MSE logct Mean
0.482954 17.14065 0.729083 4.253530

Source DF Type III SS Mean Square F Value Pr> F
x 1 1.98605237 1.98605237 3.74 0.1254

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.531562
Critical Value of t 2.77645
Least Significant Difference 1.6528

t Grouping Mean N x
A 4.8289 3 combo
A
A 3.6782 3 wounded

--------------------- ---------day=2------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 2 combo wounded

Number of Observations Read 6
Number of Observations Used 6


Dependent Variable: logct











Source
Model
Error
Corrected Total


Sum of
Squares
1.43025008
1.22623808
2.65648816


Mean Square
1.43025008
0.30655952


FValue Pr>F
4.67 0.0969


R-Square Coeff Var Root MSE
0.538399 10.98203 0.553678


DF Type III SS
1 1.43025008


Mean Square
1.43025008


FValue Pr>F
4.67 0.0969


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.30656
Critical Value of t 2.77645
Least Significant Difference 1.2552

t Grouping Mean N x
A 5.5299 3 wounded
A
A 4.5534 3 combo

--------------------- --------- day=3------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 2 combo wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total


R-Square
0.667836


Sum of
Squares
1.37074825
0.68177465
2.05252290


Coeff Var
7.432287


Mean Square
1.37074825
0.17044366


Root MSE
0.412848


FValue Pr>F
8.04 0.0471


logct Mean
5.554794


DF Type III SS
1 1.37074825


Mean Square F Value Pr > F
1.37074825 8.04 0.0471


Source
x


logct Mean
5.041675


Source











t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.170444
Critical Value of t 2.77645
Least Significant Difference 0.9359


t Grouping
A
B


Mean
6.0328
5.0768


N x
3 wounded
3 combo
Salmonella 90% RH 270C


------------------------------ day=0 --------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 2 combo wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total


R-Square
0.528231


Sum of
Squares
0.51312168
0.45827522
0.97139691


Coeff Var
15.51868


Mean Square
0.51312168
0.07637920


Root MSE
0.276368


FValue Pr>F
6.72 0.0411


logct Mean
1.780872


DF Type III SS Mean Square F Value Pr>F
1 0.51312168 0.51312168 6.72 0.0411

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 6
Error Mean Square 0.076379
Critical Value of t 2.44691
Least Significant Difference 0.4939
Harmonic Mean of Cell Sizes 3.75


t Grouping
A
B


Mean
1.9770
1.4539


N x
5 wounded
3 combo


Source
x











----------------------------------- day=1 ------------------------------------
The GLM Procedure
Class Level Information


Class
day
x


Levels Values
1 1
2 combo wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
Squares
2.05174567
4.72285687
6.77460254


Mean Square
2.05174567
0.59035711


FValue Pr>F
3.48 0.0993


Root MSE logct Mean
0.768347 4.727392


Source
x


DF Type III SS
1 2.05174567


Mean Square F Value Pr > F
2.05174567 3.48 0.0993


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 0.590357
Critical Value oft 2.30600
Least Significant Difference 1.1206


t Grouping Mean
A 5.1804


N x
5 combo


A
A 4.2744 5 wounded


------------------------------- day=2 --------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 2 combo wounded


Number of Observations Read
Number of Observations Used


R-Square
0.302858


Coeff Var
16.25309











Source
Model
Error
Corrected Total


Sum of
Squares
0.30312206
18.66443911
18.96756116


Mean Square
0.30312206
2.33305489


FValue Pr>F
0.13 0.7278


R-Square Coeff Var Root MSE
0.015981 28.44271 1.527434


DF Type III SS
1 0.30312206


Mean Square
0.30312206


FValue Pr>F
0.13 0.7278


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 2.333055
Critical Value oft 2.30600
Least Significant Difference 2.2277

t Grouping Mean N x
A 5.5443 5 wounded
A
A 5.1961 5 combo

------------------------------ day=3------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 2 combo wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total


R-Square
0.254515


Sum of
Squares
2.08961791
6.12057870
8.21019661


Coeff Var
15.13173


Mean Square
2.08961791
0.76507234


Root MSE
0.874684


FValue Pr>F
2.73 0.1370


logct Mean
5.780462


Source
x


logct Mean
5.370213






69


Source DF Type III SS Mean Square F Value Pr> F
x 1 2.08961791 2.08961791 2.73 0.1370

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 0.765072
Critical Value oft 2.30600
Least Significant Difference 1.2757

t Grouping Mean N x
A 6.2376 5 combo
A
A 5.3233 5 wounded















APPENDIX B
SHIGELLA STATISTICS

Shigella 60% RH, 270C

------------------------------ day=0 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 2 combo wounded

Number of Observations Read 10
Number of Observations Used 10
Dependent Variable: logct

Sum of
Source DF Squares Mean Square F Value Pr > F
Model 1 3.52569825 3.52569825 161.62 <.0001
Error 8 0.17452325 0.02181541
Corrected Total 9 3.70022149

R-Square Coeff Var Root MSE logct Mean
0.952834 24.87477 0.147700 0.593776

Source DF Type III SS Mean Square F Value Pr> F
x 1 3.52569825 3.52569825 161.62 <.0001

t Tests (LSD) for logct

NOTE: This test controls the Type I comparisonwise error rate, not the
experimentwise error rate.

Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 0.021815
Critical Value oft 2.30600
Least Significant Difference 0.2154

Means with the same letter are not significantly different.









t Grouping Mean N x
A 1.18755 5 combo
B 0.00000 5 wounded

------------------------------ day= ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 2 combo wounded

Number of Observations Read 10
Number of Observations Used 10

Dependent Variable: logct
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 1 0 0
Error 8 0 0
Corrected Total 9 0

R-Square Coeff Var Root MSE logct Mean
0.000000 0 0

Source DF Type III SS Mean Square F Value Pr> F
x 1 0 0

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 0
Critical Value oft 2.30600
Least Significant Difference 0

t Grouping Mean N x
A 0 5 combo
A
A 0 5 wounded

--------------------- ---------day=2------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 2 combo wounded









Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total

R-Square
0.000000


Source
x


Sum of
Squares Mean Square F Value Pr > F
0 0
0 0
0


Coeff Var Root MSE logct Mean
0 0


DF Type III SS Mean Square F Value Pr>F
1 0 0


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom
Error Mean Square 0
Critical Value oft 2.30600
Least Significant Difference


t Grouping
A


Mean N x
0 5 combo


A
A 0 5 wounded


----------------------------------- day=3 --------------------------------
The GLM Procedure
Class Level Information


Class
day
x


Levels Values
1 3
2 combo wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
1 64.56156527
8 2.23807082
9 66.79963610


Mean Square F Value Pr > F
64.56156527 230.78 <.0001
0.27975885









R-Square Coeff Var
0.966496 20.81636


Root MSE logct Mean
0.528922 2.540897


DF Type III SS Mean Square F Value Pr>F
1 64.56156527 64.56156527 230.78 <.0001

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 0.279759
Critical Value oft 2.30600
Least Significant Difference 0.7714


t Grouping
A
B


Mean N x
5.0818 5 combo
0.0000 5 wounded
Shigella 90% RH, 200C


------------------------------- day=0 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 2 combo wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total
R-Square
0.728817


Source
x


Sum of
DF Squares Mean Square F Value Pr > F
1 2.49520030 2.49520030 10.75 0.0305
4 0.92843045 0.23210761
5 3.42363076
Coeff Var Root MSE logct Mean
48.56568 0.481775 0.992008


DF Type III SS
1 2.49520030


Mean Square F Value Pr > F
2.49520030 10.75 0.0305


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.232108
Critical Value of t 2.77645
Least Significant Difference 1.0922


Source
x









t Grouping Mean N x
A 1.6369 3 wounded
B 0.3471 3 combo

------------------------------ day= ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 2 combo wounded

Number of Observations Read 6
Number of Observations Used 6

Dependent Variable: logct

Sum of
Source DF Squares Mean Square F Value Pr > F
Model 1 0.00101221 0.00101221 0.06 0.8241
Error 4 0.07191014 0.01797753
Corrected Total 5 0.07292235

R-Square Coeff Var Root MSE logct Mean
0.013881 5.099104 0.134080 2.629488

Source DF Type III SS Mean Square F Value Pr> F
x 1 0.00101221 0.00101221 0.06 0.8241

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.017978
Critical Value of t 2.77645
Least Significant Difference 0.304

t Grouping Mean N x
A 2.6425 3 combo
A
A 2.6165 3 wounded

--------------------- ---------day=2------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 2 combo wounded











Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
Squares
0.03769405
0.09492378
0.13261782


Mean Square
0.03769405
0.02373094


FValue Pr>F
1.59 0.2761


R-Square Coeff Var
0.284231 5.098507


Root MSE logct Mean
0.154049 3.021444


DF Type III SS
1 0.03769405


Mean Square
0.03769405


FValue Pr>F
1.59 0.2761


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.023731
Critical Value of t 2.77645
Least Significant Difference 0.3492

t Grouping Mean N x
A 3.1007 3 wounded
A
A 2.9422 3 combo


-------------------------------day=3------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 2 combo wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
1 0.13916497
4 1.06430764
5 1.20347261


Mean Square F Value Pr > F
0.13916497 0.52 0.5096
0.26607691


Source
x









R-Square Coeff Var Ro
0.115636 13.45912 0.5

DF Type III SS
1 0.13916497


ot MSE logct Mean
15826 3.832543

Mean Square F Value Pr > F
0.13916497 0.52 0.5096


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 4
Error Mean Square 0.266077
Critical Value of t 2.77645
Least Significant Difference 1.1694


t Grouping Mean
A 3.9848


N x
3 wounded


A
A 3.6802 3 combo
Shigella 90% RH, 270C


------------------------------- day=0 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 2 combo wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


R-Square
0.640508


Sum of
Squares
1.87514338
1.05244211
2.92758549


Coeff Var
31.74568


Mean Square
1.87514338
0.17540702


Root MSE
0.418816


FValue Pr>F
10.69 0.0170


logct Mean
1.319286


DF Type III SS Mean Square F Value Pr>F
1 1.87514338 1.87514338 10.69 0.0170


Source
x


Source









t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 6
Error Mean Square 0.175407
Critical Value of t 2.44691
Least Significant Difference 0.7484
Harmonic Mean of Cell Sizes 3.75


t Grouping
A
B


Mean
1.6943
0.6943


5 wounded
3 combo


-------------------------------day= ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 2 combo wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
1 16.55253730
8 8.50111961
9 25.05365691


Mean Square F Value Pr > F
16.55253730 15.58 0.0043
1.06263995


R-Square Coeff Var Root MSE logct Mean
0.660683 46.69006 1.030844 2.207845


DF Type III SS
1 16.55253730


Mean Square F Value Pr > F
16.55253730 15.58 0.0043


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 1.06264
Critical Value oft 2.30600
Least Significant Difference 1.5034


t Grouping
A
B


Mean
3.4944
0.9213


N x
5 wounded
5 combo


Source
x









-------------------- ---------- day=2 --------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 2 combo wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total
R-Square
0.004333


Sum of
DF Squares
1 0.12107542
8 27.82375321
9 27.94482864
Coeff Var R(
60.00267 1.


Mean Square
0.12107542
3.47796915


FValue Pr>F
0.03 0.8566


)ot MSE logct Mean
864931 3.108081


DF Type III SS
1 0.12107542


Mean Square
0.12107542


FValue Pr>F
0.03 0.8566


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 3.477969
Critical Value oft 2.30600
Least Significant Difference 2.7199

t Grouping Mean N x
A 3.218 5 combo
A
A 2.998 5 wounded


------------------------------ day=3 ----------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 2 combo wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
x











Source
Model
Error
Corrected Total


Sum of
Squares
16.05186754
9.23255826
25.28442580


Mean Square
16.05186754
1.15406978


FValue Pr>F
13.91 0.0058


R-Square Coeff Var Root MSE logct Mean
0.634852 25.32880 1.074276 4.241323

DF Type III SS Mean Square F Value Pr>F
1 16.05186754 16.05186754 13.91 0.0058

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 1.15407
Critical Value oft 2.30600
Least Significant Difference 1.5668


t Grouping
A
B


Mean
5.5083
2.9744


N x
5 combo
5 wounded


Source
x















APPENDIX C
ERWINIA STATISTICS

Erwinia 60% RH, 270C


------------------------------- day=0 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
2 10.50799005
12 3.30449643
14 13.81248648


Mean Square
5.25399502
0.27537470


FValue Pr>F
19.08 0.0002


R-Square Coeff Var
0.760760 36.37229


DF Type III SS
2 10.50799005


Mean Square F Value Pr > F
5.25399502 19.08 0.0002


t Tests (LSD) for logct

NOTE: This test controls the Type I comparisonwise error rate, not the
experimentwise error rate.

Alpha 0.05
Error Degrees of Freedom 12
Error Mean Square 0.275375
Critical Value oft 2.17881
Least Significant Difference 0.7231

Means with the same letter are not significantly different.


Source
x


Root MSE
0.524762


logct Mean
1.442751










t Grouping Mean
A 2.1293


1.9345
0.2644


N x
5 combosal


comboshi
wounded


-------------------------------day= ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used


Source
Model
Error
Corrected Total


Sum of
Squares Mean Square
1.65440723 0.82720361
12.43466317 1.03622193
14.08907040


FValue Pr>F
0.80 0.4726


R-Square Coeff Var Root MSE logct Mean
0.117425 46.16068 1.017950 2.205231


DF Type III SS
2 1.65440723


Mean Square F Value Pr > F
0.82720361 0.80 0.4726


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 12
Error Mean Square 1.036222
Critical Value oft 2.17881
Least Significant Difference 1.4027


t Grouping Mean
A 2.6748


x
combosal


A 1.9767 5 wounded
A
A 1.9641 5 comboshi


Source
x









------------------------------- day=2 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used


Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
2 48.63180315
12 33.90544483
14 82.53724798


Mean Square
24.31590157
2.82545374


FValue Pr>F
8.61 0.0048


R-Square Coeff Var Root MSE logct Mean
0.589210 48.30976 1.680909 3.479439


DF Type I SS Mean Square
2 48.63180315 24.31590157

Alpha 0.05
Error Degrees of Freedom 12
Error Mean Square 2.825454
Critical Value oft 2.17881
Least Significant Difference 2.3163


FValue Pr>F
8.61 0.0048


t Grouping
A
B


Mean
6.000
2.531


N x
5 combosal
5 comboshi


B
B 1.907 5 wounded


-------------------------------day=3------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used


Source
x









Dependent Variable: logct
Sum of
Source DF Squares Mean Square
Model 2 36.41360097 18.20680049
Error 12 29.32981256 2.44415105
Corrected Total 14 65.74341353


R-Square Coeff Var Root MSE
0.553875 30.94536 1.563378


FValue Pr>F
7.45 0.0079


logct Mean
5.052060


DF Type III SS
2 36.41360097


Mean Square
18.20680049


FValue Pr>F
7.45 0.0079


Alpha 0.05
Error Degrees of Freedom 12
Error Mean Square 2.444151
Critical Value oft 2.17881
Least Significant Difference 2.1543


t Grouping Mean
A 6.2014


6.1055
2.8493


N x
5 comboshi

5 combosal
5 wounded
Erwinia 90% RH, 200C


------------------------------- day=0 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
2 0.14695946
6 1.30692086
8 1.45388032


Mean Square
0.07347973
0.21782014


FValue Pr>F
0.34 0.7264


R-Square Coeff Var
0.101081 24.06714


Source
x


Root MSE
0.466712


logct Mean
1.939208









Source DF Type I SS Mean Square F Value Pr> F
x 2 0.14695946 0.07347973 0.34 0.7264

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 6
Error Mean Square 0.21782
Critical Value of t 2.44691
Least Significant Difference 0.9324

t Grouping Mean N x
A 2.1150 3 wounded
A
A 1.8876 3 combosal
A
A 1.8151 3 comboshi

------------------------------ day= ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 3 combosal comboshi wounded

Number of Observations Read 9
Number of Observations Used 9
Dependent Variable: logct

Sum of
Source DF Squares Mean Square F Value Pr > F
Model 2 16.51690575 8.25845287 21.09 0.0019
Error 6 2.34960666 0.39160111
Corrected Total 8 18.86651240

R-Square Coeff Var Root MSE logct Mean
0.875462 13.04686 0.625780 4.796407

Source DF Type III SS Mean Square F Value Pr> F
x 2 16.51690575 8.25845287 21.09 0.0019

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 6
Error Mean Square 0.391601
Critical Value of t 2.44691
Least Significant Difference 1.2502










t Grouping Mean
A 6.0351


5.4428
2.9113


N x
3 wounded


combosal
comboshi


-------------------------------day=2------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
2 0.80993932
6 0.99978794
8 1.80972725


Mean Square F Value Pr > F
0.40496966 2.43 0.1686
0.16663132


R-Square Coeff Var Root MSE logct Mean
0.447548 6.146878 0.408205 6.640851


DF Type I SS
2 0.80993932


Mean Square F Value Pr > F
0.40496966 2.43 0.1686


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 6
Error Mean Square 0.166631
Critical Value of t 2.44691
Least Significant Difference 0.8156


t Grouping
A


Mean
7.0541


x
wounded


A
A 6.5173 3 comboshi
A
A 6.3512 3 combosal


Source
x










--------------------- --------- day=3 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 3 combosal comboshi wounded

Number of Observations Read 9
Number of Observations Used 9
Dependent Variable: logct

Sum of
Source DF Squares Mean Square F Value Pr > F
Model 2 0.75602073 0.37801037 25.93 0.0011
Error 6 0.08746506 0.01457751
Corrected Total 8 0.84348579

R-Square Coeff Var Root MSE logct Mean
0.896305 1.580899 0.120737 7.637260

Source DF Type I SS Mean Square F Value Pr> F
x 2 0.75602073 0.37801037 25.93 0.0011

t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 6
Error Mean Square 0.014578
Critical Value of t 2.44691
Least Significant Difference 0.2412

t Grouping Mean N x
A 8.03550 3 wounded
B 7.52213 3 comboshi
B
B 7.35415 3 combosal
Erwinia 90% RH, 270C

------------------------------ day=0 ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 0
x 3 combosal comboshi wounded
Number of Observations Read 15
Number of Observations Used 11











Source
Model
Error
Corrected Total


Sum of
Squares
2.75162122
2.04224723
4.79386845


Mean Square
1.37581061
0.25528090


F Value Pr>F
5.39 0.0329


R-Square Coeff Var Root MSE
0.573988 29.09062 0.505253

DF Type I SS Mean
2 2.75162122 1.375r


logct Mean
1.736825


Square F Value Pr > F
81061 5.39 0.0329


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 8
Error Mean Square 0.255281
Critical Value oft 2.30600

Comparisons significant at the 0.05 level are indicated by ***
Difference


Between


95% Confidence


Comparison Means Limits
wounded combosal 0.5524 -0.2985 1.4032
wounded comboshi 1.2066 0.3557 2.0575 ***
combosal wounded -0.5524 -1.4032 0.2985
combosal comboshi 0.6543 -0.2971 1.6056
comboshi wounded -1.2066 -2.0575 -0.3557 ***
comboshi combosal -0.6543 -1.6056 0.2971
--------------------- --------- day= ------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 1
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used
Dependent Variable: logct


Source
Model
Error
Corrected Total


Sum of
DF Squares
2 10.32400292
12 5.71129079
14 16.03529371


Mean Square
5.16200146
0.47594090


FValue Pr>F
10.85 0.0020


R-Square Coeff Var Root MSE
0.643830 13.31123 0.689885


Source
x


logct Mean
5.182726









DF Type III SS
2 10.32400292


Mean Square F Value Pr > F
5.16200146 10.85 0.0020


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 12
Error Mean Square 0.475941
Critical Value oft 2.17881
Least Significant Difference 0.9507


t Grouping Mean
A 5.9190


5.6056
4.0235


N x
5 wounded

5 combosal
5 comboshi


-------------------------------day=2------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 2
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used


Source
Model
Error
Corrected Total


Sum of
DF Squares
2 9.31016860
12 9.49263861
14 18.80280722


Mean Square F Value Pr > F
4.65508430 5.88 0.0166
0.79105322


R-Square Coeff Var Root MSE logct Mean
0.495148 13.03674 0.889412 6.822349


DF Type III SS
2 9.31016860


Mean Square F Value Pr > F
4.65508430 5.88 0.0166


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 12
Error Mean Square 0.791053
Critical Value oft 2.17881
Least Significant Difference 1.2256


Source
x


Source
x










t Grouping
A
B


Mean
7.9329
6.3450


x
wounded
combosal


B
B 6.1892 5 comboshi


-------------------------------day=3------------------------------
The GLM Procedure
Class Level Information
Class Levels Values
day 1 3
x 3 combosal comboshi wounded


Number of Observations Read
Number of Observations Used


Source
Model
Error
Corrected Total


Sum of
DF Squares
2 12.33484301
12 2.13991544
14 14.47475845


Mean Square
6.16742151
0.17832629


FValue Pr>F
34.59 <.0001


R-Square Coeff Var Root MSE logct Mean
0.852162 5.521948 0.422287 7.647427


DF Type III SS
2 12.33484301


Mean Square F Value Pr > F
6.16742151 34.59 <.0001


t Tests (LSD) for logct
Alpha 0.05
Error Degrees of Freedom 12
Error Mean Square 0.178326
Critical Value oft 2.17881
Least Significant Difference 0.5819


t Grouping
A
B
C


Mean
8.8046
7.5475
6.5901


x
wounded
comboshi
combosal


Source
x
















LIST OF REFERENCES


Allen, R.L. 2003. A Recovery Study of Salmonella spp. from the Surfaces of Tomatoes
and Packing Line Materials; Master's Thesis [electronic resource]. Gainesville, FL.
University of Florida.

Altekurse, S.F., M.L. Cohen and D.L. Swerdlow. 1997. Emerging Foodborne Diseases.
Emerg. Inf. Dis. 3(3):285-293.

Asplund, K. and E. Nurmi. 1991. The Growth of Salmonella in Tomatoes. Int. J. Food
Micro. 13:177-182.

Aysan, Y., A. Karatas and O. Cinar. 2003. Biological Control of Bacterial Stem Rot
Caused by Erwinia c h1,\u1thei1ni on Tomato. Crop Prot. 22:807-811.

Bagamboula, C.F., M. Uyttendaele and J. Debevere. 2002. Acid Tolerance of.\l/nge/A
sonnei and .\rgell/,iflexneri. J. App. Microbiol. 93(3):479-486.

Baker, R.C., R.A. Qureshi and J.H. Hotchkiss. 1986. Effect of Elevated Level of Carbon
Dioxide on the Growth of Spoilage and Pathogenic Bacteria. Poult. Sci. 65:729-
737.

Barras, F., F. van Gijsegem and A.K. Chatterjee. 1994. Extracellular Enzymes and
Pathogenesis of the Soft Rot Erwinia. Ann. Rev. Phytopathol. 32:201-234.

Bartz, J.A. and R.K. Showalter. 1981. Infiltration of Tomatoes by Aqueous Bacterial
Suspension. Phytopath. 71:515-518.

Beattie, G.A. and S.E. Lindow. 1999. Bacterial colonization of leaves: A spectrum of
strategies. Phytopathol. 89(5):353-9.

Beuchat, L.R. 1995. Pathogenic Microorganisms Associated with Fresh Produce. J. Food
Prot. 59(2):204-216.

Beuchat, L.R. 2002. Ecological Factors Influencing Survival and Growth of Human
Pathogens on Raw Fruits and Vegetables. Microbes Infect. 4:413-423.

Beuchat, L.R., A.J. Scouten, R.I. Allen and R.S. Hussey. 2003. Potential of a Plant-
Parasitic Nematode to Facilitate Internal Contamination of Tomato Plants by
Salmonella. J. Food Prot. 66(8):1459-1461.




Full Text

PAGE 1

SURVIVAL OF Salmonella AND Shigella ON TOMATOES IN THE PRESENCE OF THE SOFT ROT PATHOGEN, Erwinia Carotovora By JENNIFER A. JOY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Jennifer A. Joy

PAGE 3

To my parents, John and Dolores Joy; th eir unwavering love and support made this possible

PAGE 4

ACKNOWLEDGMENTS I would like to thank my supervisory committee chair (Dr. Keith R. Schneider) for his assistance and guidance. My supervisory committee members (Dr. Douglas L. Archer and Dr. Jerry A. Bartz) are also due many thanks for their help in my research. Six Ls Packing Company, Inc. and DiMare Fresh, Inc. are greatly appreciated for always providing top-quality tomatoes. Without these produce suppliers, much of my research would not have been completed. I would like to extend thanks to all of my lab mates for all of their hard work and diligence in assisting me with research projects. This research was funded in part by the USDA-CSREES (Grant number 00-52102-9637), responsible for providing the .33 FTE research assistantship. Additionally, I would like to thank the Institute of Food and Agricultural Sciences (IFAS) Statistics Department for all of their help in analyzing my results. Most importantly, I would like to thank my parents, John and Dolores Joy, for all of their love and encouragement. My achievements of the past 2 years would not have been possible without them. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................4 Importance of Foodborne Illness Related to Fresh Produce.........................................4 Foodborne Illness Linked to Produce...........................................................................5 Salmonella.............................................................................................................7 Salmonellosis.........................................................................................................9 Shigella................................................................................................................10 Shigellosis............................................................................................................11 Erwinia carotovora.............................................................................................12 Tomatoes.............................................................................................................14 Tomato Market and Regulations.........................................................................15 Tomatoes and Pathogens.....................................................................................16 Quorum Sensing..................................................................................................17 3 MATERIALS AND METHODS...............................................................................20 Selection of Temperature and Relative Humidity Conditions....................................20 Acquisition and Maintenance of Salmonella Cultures...............................................21 Acquisition and Maintenance of Shigella Cultures....................................................22 Acquisition and Maintenance of Erwinia carotovora Culture...................................23 Growth Studies...........................................................................................................23 Preparation of Inoculum.............................................................................................24 Inoculation of Tomatoes.............................................................................................25 Intact Tomatoes...................................................................................................25 Shave-Wounded Tomatoes..................................................................................26 v

PAGE 6

Pathogen Recovery off Tomato Surfaces...................................................................26 Intact Tomatoes...................................................................................................26 Shave-Wounded Tomatoes..................................................................................27 Shave-Wounded Tomatoes with Combined Pathogens......................................28 4 RESULTS...................................................................................................................30 Growth Levels of Salmonella, Shigella and Erwinia carotovora...............................31 Preliminary Recovery Studies: Intact Tomato Surfaces.............................................33 Recovery of Bacteria from Tomato Surfaces.............................................................33 Recovery of Bacteria from Intact Tomato Surfaces...................................................34 Salmonella Recovery off Intact Tomato Surfaces...............................................34 Shigella Recovery from Intact Tomato Surfaces.................................................35 Erwinia Recovery from Intact Tomato Surfaces.................................................36 Recovery of Bacteria from Shave-Wounded Tomato Surfaces..................................38 Salmonella Recovery from Shave-Wounded Tomato Surfaces..........................38 Shigella Recovery from Shave-Wounded Tomato Surfaces...............................39 Erwinia Recovery from Shave-Wounded Tomato Surfaces...............................40 Recovery of Combined Bacteria from Shave-Wounded Tomato Surfaces................41 Recovery of Salmonella and Erwinia in Fall/Winter Season Conditions...........41 Recovery of Salmonella and Erwinia in Optimum Conditions for Erwinia.......42 Recovery of Salmonella and Erwinia in Standard Ripening Room Conditions.43 Recovery of Shigella and Erwinia in Fall/Winter Season Conditions................44 Recovery of Shigella and Erwinia under Optimum Conditions for Erwinia......46 Recovery of Shigella and Erwinia in Standard Ripening Room Conditions......47 5 DISCUSSION.............................................................................................................48 Recovery of Bacteria from Intact Tomato Surfaces...................................................50 Recovery of Bacteria from Wounded Tomato Surfaces.............................................53 Recovery of Combined Bacteria from Wounded Tomato Surfaces...........................55 6 CONCLUSION...........................................................................................................58 APPENDIX A SALMONELLA STATISTICS....................................................................................60 B SHIGELLA STATISTICS...........................................................................................70 C ERWINIA STATISTICS.............................................................................................80 LIST OF REFERENCES...................................................................................................90 BIOGRAPHICAL SKETCH.............................................................................................96 vi

PAGE 7

LIST OF TABLES Table page 2-1 Shigella species designated by their serogroup and number of serotypes...............11 2-2 Tomato size designation...........................................................................................16 3-1 Temperature and relative humidity conditions selected to simulate a standard ripening room environment (90% RH, 20C) and optimum temperature for the growth of Erwinia at fall/winter conditions (60% RH, 27C) and a ripening room (90% RH, 27C)..............................................................................................21 3-2 Salmonella enteritidis serovars obtained from Dr. Linda J. Harris at the University of California, Davis: wild type serovars listed with source....................22 4-1 The log10 CFU/mL reduction of individual bacterias off intact tomato surfaces during preliminary studies........................................................................................33 4-2 Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in fall/winter (60% RH, 27C) conditions over three days..........................................42 4-3 Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in optimized conditions for Erwinia carotovora (90% RH, 27C) over three days.....43 4-4 Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in standard ripening room conditions (90% RH, 20C) over three days......................44 4-5 Recovered Shigella and Erwinia from shave-wounded tomato surfaces in fall/winter season conditions (60% RH, 27C) over three days...............................45 4-6 Recovered Shigella and Erwinia from shave-wounded tomato surfaces under optimum conditions for Erwinia carotovora (90% RH, 27C) over three days......46 4-7 Recovered Shigella and Erwinia from shave-wounded tomato surfaces under standard ripening room conditions (90% RH, 20C) over three days......................47 vii

PAGE 8

LIST OF FIGURES Figure page 4-1 Average log10 CFU/mL growth curves of Salmonella serovars over a 10-hour incubation.................................................................................................................31 4-2 Average log10 CFU/mL growth curves of Shigella species over a 12-hour incubation.................................................................................................................31 4-3 Average log10 CFU/mL growth curves of Erwinia carotovora over a 21-hour incubation.................................................................................................................32 4-4 Maximum average log10 CFU/mL counts of Salmonella serovars, Shigella species and Erwinia (rif+) after specific incubation times.......................................32 4-5 Salmonella recovery (log10 CFU/mL) from intact tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days..................................................35 4-6 Shigella recovery (log10 CFU/mL) from intact tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days..................................................36 4-7 Erwinia carotovora recovery (log10 CFU/mL) from intact tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days.............................37 4-8 Salmonella recovery (log10 CFU/mL) from shave-wounded tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days.............................38 4-9 Shigella recovery (log10 CFU/mL) from shave-wounded tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days.............................40 4-10 Erwinia carotovora recovery (log10 CFU/mL) from shave-wounded tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days..........41 viii

PAGE 9

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SURVIVAL OF Salmonella AND Shigella ON TOMATOES IN THE PRESENCE OF THE SOFT ROT PATHOGEN, Erwinia Carotovora By JENNIFER JOY December 2005 Chair: Keith R. Schneider Major Department: Food Science and Human Nutrition Recently, outbreaks of Salmonella and Shigella have been associated with the consumption of fresh produce. Previous studies have investigated the survival of Salmonella and Shigella on fresh produce, however there is little data on the effects of bacterial soft rot on human pathogen survival. Erwinia carotovora is the bacterial species most commonly associated with soft rot on fresh vegetables. My study investigated the survival of Salmonella and Shigella with and without the presence of Erwinia carotovora on intact and compromised tomato surfaces. Whole, unripe tomatoes were artificially inoculated with either a 5-strain Salmonella cocktail or a 2-strain Shigella cocktail (with and without Erwinia carotovora). Inoculated tomatoes were incubated at room temperature (~27C) under controlled relative humidity to simulate fall/winter in Florida (60%) and optimum growth conditions for Erwinia carotovora (90%). Additionally, inoculated tomatoes were incubated at standard ripening-room conditions of 90% relative humidity and 20C. ix

PAGE 10

All three environmental conditions allowed for the survival of Salmonella on intact tomato surfaces. Significant population decreases were observed, but viable cells were still present after 3 days. Optimum growth conditions for Erwinia carotovora allowed the best survival of Shigella and Erwinia on intact tomato surfaces. Wounded tomato surfaces allowed growth of the bacteria except for Shigella at fall/winter season conditions. No viable cells were recovered over the span of 3 days. Salmonella and Shigella on wounded surfaces reached peak growth at standard ripening room conditions (90% RH, 20C), while Erwinia optimum conditions for growth were 90% RH, 27C. In combined bacteria studies on wounded tomato surfaces using Salmonella and Erwinia, peak growth of Salmonella occurred under optimum conditions for growth of Erwinia (4.81 log10 CFU/mL), while peak growth of Erwinia occurred under standard ripening-room conditions (6.26 log10 CFU/mL). For studies combining Shigella and Erwinia, peak growth occurred for both microorganisms when held under optimum growth conditions for Erwinia (4.84 log10 CFU/mL and 6.55 log10 CFU/mL respectively). My results suggest that temperature and relative humidity (not temperature alone) play an important role in the growth and survival of bacteria on both intact and wounded tomato surfaces. The risk of foodborne illness from fresh produce can be lessened by ensuring good culling practices and establishing a lower humidity in packinghouses to prevent the growth of bacteria on produce, specifically tomatoes. x

PAGE 11

CHAPTER 1 INTRODUCTION In recent years, increases in the number of outbreaks of human illness have been associated with consumption of fresh fruits and vegetables (Tauxe et al. 1997). Several factors may have contributed to this increase including the emergence of new pathogens, adaptation of pathogens to environmental stresses imposed by fruits and vegetables, increased global travel facilitating exposure to pathogens from other geographical areas, and agricultural practices that use improperly treated irrigation water or animal manure (Beuchat 1995; Tauxe et al. 1997; Wade and Beuchat 2003). Salmonella serotypes cause approximately 1.5 million cases of foodborne gastroenteritis annually in the U.S., resulting in approximately 15,000 hospitalizations and 500 deaths (Mead et al. 1999). Shigella spp. have a relatively low infectious dose, requiring as few as 10 to 1000 cells to cause illness (Smith 1987). Yearly in the U.S., Shigella spp. are responsible for approximately 450,000 cases of foodborne illness, resulting in 1,300 hospitalizations and 14 deaths (Mead et al. 1999). The Centers for Disease Control and Prevention (CDC) report that the number of produce-associated outbreaks and cases of foodborne illness has more than doubled between the periods of 1973 to 1987 and 1988 to 1991 (Tauxe et al. 1997). In 1999, the FDA conducted a survey of produce commonly consumed in the U.S., including tomatoes, cantaloupe, green onions, strawberries, and celery. They tested a total of 1003 samples for the presence of Salmonella, Shigella, or E. coli O157:H7. Of the samples analyzed, 4% were found to be contaminated with either Salmonella or Shigella (FDA 2001). In response to 1

PAGE 12

2 that study, the FDA repeated this produce survey in 2000 and determined that 1% of the samples was contaminated with either Salmonella or Shigella and none were found to be contaminated with E. coli O157:H7 (FDA 2003). High-acid fruits and vegetables (pH less than 4.6), previously were considered unable to support the growth of bacteria that cause human infections (Bagamboula et al. 2002). Some foodborne pathogens have shown surprising tolerance to stressful environments. Salmonella serovars have been shown to grow at temperatures ranging from 2C (Baker et al. 1986) to 54C (Droffner and Yamamoto 1992) and over a pH range from 3.99 (Asplund and Nurmi 1991) to 9.5 (Holley and Proulx 1986). Several recent outbreaks of salmonellosis in the U.S. have been associated with consumption of uncooked tomatoes, including S. enterica serotypes Javiana and Montevideo (Hedberg et al. 1999). Additionally, raw tomatoes have been implicated in outbreaks involving S. enterica serotype Baildon (Cummings et al. 2001). The reported occurrence of foodborne shigellosis, primarily caused by Shigella sonnei and Shigella flexneri, is lower than that of salmonellosis or other enteric pathogens. Each year there are a significant number of outbreaks of shigellosis, many associated with consumption of fresh produce (Smith 1987; Mead et al. 1999). The most important contributing factor leading to Shigella foodborne outbreaks is poor personal hygiene of a food handler (Smith 1987). Fruits and their products should be handled properly to prevent contamination. The increase in worldwide food trade and fresh produce consumption, and the popularity of minimally processed fruits and vegetables may have a significant role in introducing Shigella spp. to these types of foods (Bagamboula et al. 2002).

PAGE 13

3 Most of the bacteria responsible for spoilage of fresh-cut fruits and vegetables are Gram-negative. Erwinia carotovora is the species most commonly associated with decay of vegetables (Lund 1983) such as potato, tomato, onion, corn, rice, and sugar beets (Aysan et al. 2003). Soft rot Erwinias cause tissue maceration (referred to as soft rot disease), by producing cell-wall-degrading enzymes (Toth et al. 2003). The main weapon used by soft rot Erwinia is coordinated production of high levels of multiple exoenzymes including pectinases and proteases, which break down plant cell walls and release nutrients for bacterial growth (Barras et al. 1994). My study evaluated the survival and recovery of Salmonella spp. and Shigella spp. from tomato surfaces individually and in the presence of the soft-rot plant pathogen, Erwinia carotovora. Intact and shave-wounded tomatoes were inoculated with known concentrations of the rifampicin-resistant variants of the bacteria and combinations of Salmonella-Erwinia and Shigella-Erwinia. Bacterial recovery off of intact tomato surfaces was accomplished by a rub-shaking method (Burnett and Beuchat 2001; Harris et al. 2002; Zhuang et al. 1995), while the shave-wounded tomato samples were excised and stomached. Tomatoes were kept in an environmental humidity chamber for 3 days to maintain specific temperature and humidity conditions. Conditions were established to simulate the typical Florida fall/winter production season (60% RH, 27C), optimum conditions for the survival of Erwinia carotovora (90% RH, 27C), and standard tomato ripening rooms (90% RH, 20C).

PAGE 14

CHAPTER 2 LITERATURE REVIEW Importance of Foodborne Illness Related to Fresh Produce During the last 25 years, the number of foodborne disease outbreaks and illnesses has increased (Mead et al. 1999; Beuchat 2002). Outbreaks of foodborne illness have been attributed to the consumption of contaminated fresh vegetables and, to a lesser extent, fruits (Beuchat 1995; Brackett 1999). Despite great advances in science and technology, foodborne illness is still a major problem even in highly developed countries such as the U.S. and Canada. Additionally, there has been a significant increase in the different types of foods that have been found to harbor illness-causing foodborne bacteria (Samelis and Sofos 2003). Consumers recognize that fresh fruits and vegetables are important components of a healthy and balanced diet. In recent years, increased demand for fresh fruits and vegetables has arisen from studies showing positive correlations between diet and health. Additionally, whole and lightly processed fruits and vegetables have gained interest due to greater selection and year-round availability (Lucier et al. 2000). Contamination of fruits and vegetables by pesticides, plant and human pathogens, and other pollutants is becoming a larger concern to consumers and to the fresh-produce industry (Beuchat 1995; Brackett 1999; Wilson and Droby 2001). Salmonellosis outbreaks have been attributed to contaminated tomatoes, bean sprouts, cantaloupe, and watermelon (Beuchat 1996); and have been known to cause invasive disease or reactive arthritis (Altekruse et al. 1997). A Shigella flexneri 4

PAGE 15

5 gastroenteritis outbreak was associated with green onions (Beuchat 1996) and an outbreak of E. coli O157:H7, the enterohemorrhagic serotype, has been linked to the consumption of contaminated cantaloupes (Beuchat 1996). Foodborne Illness Linked to Produce Incidence of foodborne illness attributed to fresh produce is low, but there is always the possibility of fruits and vegetables being contaminated. Concern about contaminated fruits and vegetables is well founded because of the increase in consumption of fresh produce and the increase of produce imported from countries where standards of sanitation and handling are more lenient than here in the U.S. (Beuchat 1996). Salmonella serotypes cause approximately 1.5 million cases of foodborne gastroenteritis annually in the U.S. resulting in approximately 15,000 hospitalizations and 500 deaths (Mead et al. 1999). Wells and Butterfield (1997) showed a strong statistical association between the occurrence of bacterial soft-rot plant pathogen (Erwinia carotovora) and the presence of Salmonella on tomatoes. Of the bacterial isolates recovered from the tomatoes used in the study, 30% were confirmed to be Salmonella. Liao and Sapers (1999) found an antagonistic relationship between soft-rotting bacteria and L. monocytogenes on potato slices. Additionally, Pseudomonas fluorescens prohibited the growth of Listeria, but E. carotovora enhanced growth of this human pathogen on potato slices. It is unclear whether these responses are caused by positive or negative interactions between the pathogen and the spoilage organisms or these responses are caused by increased availability of nutrients in the macerated tissue produced as a result of soft rot (Novak et al. 2003).

PAGE 16

6 To determine the contamination rate of imported fresh produce, in March 1999, the U.S. Food and Drug Administration (FDA) conducted a survey of 1003 samples from 21 countries. The samples included broccoli, cantaloupe, celery, cilantro, loose-leaf lettuce, parsley, scallions, strawberries and tomatoes. Samples were analyzed for Salmonella, Shigella and E. coli O157:H7. Four percent (44 of 1003 samples) were contaminated with either Salmonella or Shigella. Of these contaminated samples, 80% (35 of 44 samples) were contaminated with Salmonella, and 20% (9 of 44 samples) were contaminated with Shigella. The produce with the most contamination of Salmonella and Shigella was cantaloupe and cilantro (FDA 2001). As a follow-up to these findings, in March of 2000, the FDA conducted another produce survey including 1000 samples from approximately 18 states. High volume produce such as cantaloupe, celery, cilantro, loose-leaf lettuce, parsley, scallions, strawberries and tomatoes were collected and analyzed for contamination with Salmonella, Shigella and E. coli O157:H7. Of the total 1028 domestic samples, 99% were not contaminated with any of the mentioned pathogens. Eleven samples (1% of total samples) were contaminated with either Salmonella or Shigella and no samples analyzed were contaminated with E. coli O157:H7. Of the 11 contaminated samples, six (55%) were contaminated with Salmonella and five (45%) were contaminated with Shigella, with cantaloupe and scallions resulting in the majority of positive samples (FDA 2003). Often fresh produce is washed in tap water and left to air dry. Kelman (1956) found that the viability of a soil-borne bacterium, Ralstonia solanacearum could survive in tap water for several months. In response to this research, a study by Liao and

PAGE 17

7 Shollenberger (2003) was performed to determine the suitability of using sterile water and phosphate buffered saline for preservation of bacteria that were pathogenic to plants or humans. Their findings show that soft rot bacteria including Erwinia spp. remain viable for 12-16 years in sterile distilled water, while the foodborne pathogen Salmonella was able to survive for 5 years (Liao and Shollenberger 2003). Fresh produce is also exposed to sanitizers in the processing water, primarily hypochlorite. Many consumers incorrectly assume that the sanitizers are used to treat the fruit or vegetable, when in actuality, the sanitizers are primarily used to maintain bacteriological quality of the water (Bracket 1999). In a study by Zhuang et al. (1995), the effect of sanitizers and surfactants were of minimal value in reducing populations of Salmonella in tomato fruit demonstrating that sanitizers may help but cannot guarantee the complete elimination of pathogens from produce (Brackett 1999). Salmonella The foodborne organism Salmonella was first recognized in the late 1800s. In 1885, veterinary pathologist D.E. Salmon isolated the microorganism Bacillus cholerae-suis from pigs suffering from hog cholera (Cox 2000a). Following this discovery, other similar microorganisms were isolated from foodborne outbreaks originating from food animals. To accommodate these new findings the genus Salmonella was created in honor of Salmon. The genus Salmonella is a member of the family Enterobacteriaceae and is comprised of facultative anaerobic, oxidase-negative, catalase positive, Gram-negative, rod-shaped bacteria, ranging in size from 0.7-1.5 x 2-5 m (Cox 2000a). Most Salmonella strains are motile and are capable of fermenting glucose with the production of acid and gas (FDA 2005).

PAGE 18

8 Salmonella spp. can be separated into serotypes or serovars by characterizing two major antibodies designated the O (somatic antigen) or H (flagellin antigen). The O antigen is divided into serogroups, based on differences in epitopes of the major outer membrane component of Gram-negative bacteria, the lipopolysaccharide (LPS). The H antigen is used to establish variations associated with the subunit proteins of the flagella (Cox 2000a). There are several environmental conditions that affect the growth, death or survival of Salmonella including temperature, pH and water activity (aw). The growth range for this organism is 2C (Baker et al. 1986) to 54C (Droffner and Yamamoto 1992), with the optimum being 37C (Cox 2000a). This optimum temperature is to be expected given the natural environment for most Salmonella strains of public health significance are found in the gastrointestinal tract of warm-blooded animals. Often foods high in solid content, especially protein or fat, can offer protection for the microorganism at high temperatures (Cox 2000a). Salmonella are capable of growing on defined media without special growth factors, with most strains being aerogenic (Buchanan and Gibbons 1974). The pH optimum range for growth of Salmonella is 6.5-7.5 and survival is possible up to 9.5 (Holley and Proulx 1986) and as low as 3.99 (Asplund and Nurmi 1991). An increase in temperature has been shown to increase sensitivity to low pH, as well as the presence of food additives like salt or nitrite. The optimum water activity for this microorganism is 0.995, with no growth being observed below 0.93. The survival time of the bacteria increases as the water activity decreases, allowing survival in low-moisture foods such as peanut butter and chocolate for several months (Cox 2000a).

PAGE 19

9 Salmonellosis After its discovery in 1885, Salmonella was recognized as an agent of disease and is still a major threat to animals and humans. Salmonella serotypes cause an estimated 1.5 million cases of foodborne gastroenteritis annually in the U.S. Of these cases, 15,000 hospitalizations and 500 deaths occur (Mead et al. 1999). Most people who are infected with Salmonella develop diarrhea, fever, and abdominal cramps 12 to 72 hours after infection. The illness lasts approximately 4 to 7 days, and most people recover without treatment. In some people the diarrhea may be so severe that the patient needs to be hospitalized. In these patients, the Salmonella infection may spread from the intestines to the blood stream, and then to other body sites and can cause death unless the person is treated promptly with antibiotics. The elderly, infants, and those with impaired immune systems are more likely to have a severe illness (CDC 2004a). The food products usually responsible for causing illness are eggs and poultry meat. Fresh fruits and vegetables have also been implicated in foodborne outbreaks (Yoon et al. 2004). Most recent cases of Salmonella associated with fresh fruits and vegetables are suspected to result from the improper storage or handling of prepared foods that initially carried the bacteria as surface contamination. Salmonella that are present on fruits and vegetables are capable of multiplying if specific extrinsic factors are there, such as improper refrigeration during storage and preparation, poor product quality, or the presence of bacterial soft rot (Wells and Butterfield 1999). Studies have shown that tomatoes have been contaminated with S. Javiana (CDC 2002), S. Montevideo (CDC 1993, Hedberg et al. 1994), S. Poona (CDC 1991) and S. Baildon (Cummings et al. 2001). S. Oranienberg (CDC 1979) and S. Javiana (Blostein 1991) have both been implicated in infections due to the consumption of precut watermelon, while S. Chester

PAGE 20

10 (Ries et al. 1990) and S. Poona (CDC 1991) have been found to be the cause of foodborne illness outbreaks in cantaloupe. Salmonellosis can lead to invasive disease or reactive arthritis (Altekruse et al. 1997). Shigella Shigella was named after the Japanese bacteriologist, K. Shiga, who was first to discover the dysentery bacillus (Buchanan and Gibbons 1974). Shigella is a Gram-negative, non-motile, non-sporeforming, rod-shaped bacteria (Lampel et al. 2000). Shigella is a facultative anaerobe that is capable of growing at temperatures as low as 6C and as high as 48C. The pH range for this organism is 4.8-9.3 and can survive 5-10 days in acidic foods such as orange juice and as much as 50 days in foods with a more neutral pH such as milk, flour and eggs. This bacterium is a member of the family Enterobacteriaceae and is closely related to Escherichia coli (Lampel et al. 2000). The species S. dysenteriae, S. flexneri, S. boydii and S. sonnei make up the genus Shigella and are all pathogenic to humans (Beuchat 1996). These four species have different virulence levels with Strain D, S. sonnei, being the causative agent of most cases of Shigella-related diarrhea (shigellosis) (Lampel et al. 2000). The geographic distributions and epidemiology for these species is different. S. dysenteriae, mainly found in the Indian subcontinent, is responsible for the most severe epidemics of dysentery. S. flexneri and S. sonnei are most commonly isolated in developed countries while S. boydii is rarely isolated in developed countries (Lampel et al. 2000). Humans harbor Shigella and the organism is easily spread from host to host. The source of infection is usually food or water that has been contaminated by the feces of human carriers (ICMSF 2002).

PAGE 21

11 Shigella can be classified into four groups; A, B, C, and D, based on biochemical reactions and the O antigen type. These groups are then subdivided into serotypes based on the O antigen (Table 2-1). Table 2-1. Shigella species designated by their serogroup and number of serotypes Species Serogroup Serotypes S. dysenteriae A 15 S. flexneri B 6 S. boydii C 19 S. sonnei D 1 It is estimated that 450,000 cases of shigellosis occur every year in the U.S. (Mead et al. 1999). Shigellosis is mainly a disease found in developing countries where sanitary waste treatment and water purification are not effective. Shigella has a low infectious dose, which allows for high rates of communicability between humans and in some cases subhuman primates. It has been determined that less than 100 cells of Shigella are capable of producing an illness (DuPont 1990). Consequently, Shigella is capable of easily infecting a crowded population most frequently by the fecal-oral route. Shigella is commonly found in water polluted with human feces and can be spread by flies, food and among people with unacceptable personal hygiene (Smith 1987). After being infected with Shigella, the incubation period is between 1 and 7 days with symptoms first appearing after 3 days. It is possible to be a carrier of this microorganism for up to four weeks (Lampel et al. 2000). Shigellosis Shigella species, mainly S. sonnei and S. flexneri are the cause of approximately 450,000 cases of foodborne illness in the U.S. yearly. Of these, 1,300 result in hospitalizations and 14 deaths (Mead et al. 1999). Shigella is considered an invasive pathogen and shigellosis occurs when virulent Shigella organisms attach to, and penetrate

PAGE 22

12 epithelial cells of the large intestine. Once they have invaded the tissue, the bacteria are released into the cytoplasm of the cell and multiply intracellularly. This infection is spread to the neighboring cells through projections from the originally infected cell (Lampel et al. 2000). The intestinal epithelial cells being attacked continue this cycle causing damage resulting in dysentery. These virulence factors are present on a large 220 kb virulence plasmid. The virulence proteins contained here are regulated by different environmental stimuli, with temperature being most influential. Virulence gene expression in Shigella is repressed at 30C and enabled at 37C (Lampel et al. 2000). A large S. sonnei outbreak involving 347 people was traced to commercially distributed shredded lettuce. It was determined that an infected handler was contaminating the lettuce during the shredding process (Brackett 1992). More recently, two midwestern U.S. outbreaks of S. flexneri have been associated with green onions, resulting in gastroenteritis (Beuchat 1996). Erwinia carotovora The bacterial plant pathogen Erwinia is known to cause soft rot in many crops worldwide. The predominant plant pathogens in this genus consist of Erwinia carotovora ssp. atroseptica (Eca), E. carotovora ssp. carotovora (Ecc) and E. chrysanthemi (Ech). The soft-rot Erwinia are members of the Enterobacteriaceae, along with other plant pathogens such as Erwinia amylovora and human pathogens such as E. coli, Shigella spp., Salmonella spp. and Yersinia spp (Toth et al. 2003). Ech is known to cause stem rot on tomatoes in greenhouses. It is capable of inducing wilt throughout the whole plant, which results in the entire plant collapsing. The disease first affects the roots of plants, most likely by seed-borne or soil-borne

PAGE 23

13 infection and then can spread to other plants by cultural practices. When conditions for disease are favorable, the symptoms can develop quickly (Aysan et al. 2003). Ecc mainly affects crops in subtropical and temperate regions and has probably the widest host range, including brussel sprouts, carrots, celery, cucumbers, turnips, and potatoes, although there are many other crops that are rotted by these microorganisms post-harvest (Toth et al. 2003). A pathogenic isolate of Ecc was injected into the centers of healthy cucumber fruits attached to the vine without causing disease. However, the bacterium was detected in the internal tissues of fruits harvested from the inoculated plants (Guo et al. 2001). The soft rot Erwinia are found on plant surfaces and in soil where they may enter the plant through wound sites or natural openings on the plant surface. Once inside the plant the bacteria survive in the vascular tissue and intercellular spaces where they remain inactive until environmental conditions become suitable for disease development (Prombelon and Kelman 1980; Prombelon and Salmond 1995). In addition to free water and oxygen depletion, temperature is an important factor in disease development, and can influence which of the soft rot Erwinia cause disease. For example, it was shown that a soil temperature of 20C was an important transition point, above which Eca, and below which Ech, were not apparently pathogenic (Prombelon et al. 1987a). In a study by Prombelon et al. (1987b), the abilities of the soft rot Erwinia to grow at different temperatures were clearly displayed in vitro, where it was used to differentiate the pathogens. The findings showed that at 27C all three pathogens grew; at 33.5C, only Ecc and Ech grew; at 37C only Ech grew (Prombelon et al. 1987b).

PAGE 24

14 In addition to differences in growth, a thermal regulation of the production of cell wall degrading enzymes or exoenzymes, has been demonstrated (Nguyen et al. 2002). The main weapon employed by the soft rot Erwinia is the coordinated production of high levels of multiple exoenzymes, including pectinases and proteases, which break down plant cell walls and release nutrients for bacterial growth (Barras et al. 1994; Prombelon 2002; Thomson et al. 1999). Endoglucanase activity is capable of breaking down cellulose in the primary and secondary cell walls of the host plant. Pectinases are the main exoenzymes involved in disease development. These exoenzymes break down and utilize pectins in the middle lamella and plant cell walls, causing tissue collapse, cell damage and cell leakage (Toth et al. 2003). Tomatoes Lycopersicon esculentum, commonly referred to as the tomato, is a member of the Solanaceae family (Olson et al. 2004). Tomatoes are mostly water averaging 93.76 grams per 100 grams of edible fruit. Tomatoes are high in the minerals magnesium, potassium and phosphorus. Tomatoes are also good sources of lycopene, vitamin C, vitamin A and folate (USDA 2004). One medium sized tomato provides 40% of the RDA of vitamin C (ascorbic acid), 20% of the RDA of vitamin A, significant amounts of potassium, dietary fiber, calcium, with as few as 35 calories (Sargent 1998). Ripe tomatoes are soft, bruise easily, and begin to decline in quality after only a few days. Tomatoes ripen off the vine in response to the natural ripening chemical ethylene, which is produced by the fruit (Sargent 1998). Traditionally, growers pick the fruits in the green-mature stage just as the fruit reach full size. The fruit is then shipped to other locations and can resist bruising or rotting because of its firmness. The fruits are

PAGE 25

15 usually red by the time they reach their destination, or they can be induced to ripen with the application of an ethylene gas (CDC 2004b). Tomatoes are currently one of the most popular vegetables among American consumers, with Florida producing 40% of all commercially grown fresh tomatoes (Sargent 1998). Tomatoes are members of the fruit family but are served and prepared as a vegetable. The National Cancer Institute has shown through extensive research that a diet rich in tomato-based foods can decrease the incidence of prostate cancer (CDC 2004b). Tomato fruits have a thin epidermis which makes them easily compromised by mechanical pressure, which can result in punctures, cracks, abrasions, and insect wounds that render the fruit susceptible to preharvest and postharvest microbial invasion. The stem scar tissue is also capable of absorbing water and any microorganisms that may be present (Bartz and Showalter 1981). Tomatoes should be stored at cool room temperature since storage below about 16C can damage the fruit and lead to poor quality (Sargent 1998). They should appear red or reddish-orange when ripe and should be free from bruises, blemishes or cracks (CDC 2004b). Tomato Market and Regulations The U.S. is one of the five leading tomato-producing countries. In 1985, the per capita consumption of raw tomato fruit in America was 16.6 lb, which in 1995 increased to 18.8 lb (Guo et al. 2001). This increase in consumption is due to the expansion of the domestic greenhouse industry. Tomatoes are second only to potatoes in both U.S. farm value and vegetable consumption. Over the past 20 years, U.S. annual per capita use of tomatoes and tomato products has increased nearly 30 percent, making a farm value of

PAGE 26

16 $1.8 billion (Lucier et al. 2000). The U.S. Standards for Grades of Fresh Tomatoes give size designations (Table 2-2): Table 2-2. Tomato size designation Size Designation Minimum Diameter (inches) Maximum Diameter (inches) Small 24/32 29/32 Medium 28/32 217/32 Large 216/32 225/32 Extra Large 224/32 When grading tomatoes, the following color terms may be used: Green: The color of the tomato is completely green in color. Breakers: A definite break in color from green to yellow or orange on not more than 10% of the surface. Turning: More than 10% but less than 30% in the color change from green to red. Pink: with more than 30% but less than 60% of the surface shows a pink or red color. Light red: More than 60% but less than 90% of the surface is a red color. Red: More than 90% of the surface being red in color (USDA 1997). Tomatoes and Pathogens Tomatoes have been involved in many multistate outbreaks of Salmonella (CDC 2002, Cummings et al. 2001, Hedberg et al. 1999). Several of these outbreaks can be traced back to the packinghouse where tomatoes are normally dumped into a common water bath (Hedberg et al. 1999). If the wash water or surface of the tomatoes is contaminated, bacteria can possibly contaminate the interior of the fruit resulting in faster decay. It is also possible for bacteria to survive on the fruits surface and then be transferred to the flesh during handling or cutting (Ibarra-Sanchez et al. 2004). Field-grown plants are continuously exposed to many soil-inhabiting microorganisms. Root diseases are often caused by interactions between organisms.

PAGE 27

17 Plant parasitic nematodes are capable of making plant roots more susceptible to invasion and pathogenesis by other microorganisms. In a recent study by Beuchat et al. (2003), the potential role of an endoparasitic root knot nematode, Meloidogne incognita, in facilitating the entry of Salmonella into tomato root tissues was studied. It was determined that 31 of 36 samples infested with M. incognita and Salmonella were positive for Salmonella four weeks after inoculation. In a study by Guo et al. (2001), the possibility of internalization of Salmonella in tomato fruits developed from inoculated flowers and stems was observed. Salmonella was detected in stem scar tissue and pulp of tomatoes from inoculated plants. It was also detected on or in tomatoes from plants receiving stem inoculation before or after flower set, and on or in tomatoes that developed from inoculated flowers. The highest percentage of Salmonella was found on the surface of the tomato and around stem scar tissue (Guo et al. 2001). Quorum Sensing Quorum sensing is understood to occur as a cell density dependent signaling system that occurs in many genera of bacteria (Smith et al. 2003). Initiation of a concerted population response depends on the population reaching a minimal population unit or quorum (Fray 2001). This mechanism enables bacteria to alter several cellular functions including sporulation, biofilm formation, bacteriocin production, and virulence responses among others. Quorum sensing involves celltocell communication and is controlled by extracellular chemical signals referred to as autoinducers or bacterial pheromones, produced by the bacteria when certain cell densities are reached. When these specific cell densities or signals are reached, target genes are either repressed or activated once recognized (Smith et al. 2003).

PAGE 28

18 An advantage of quorum sensing is that it enables bacteria to access favorable environments or nutrients. It also increases bacterias ability to defend themselves against eukaryotic hosts, competing bacteria, or environmental stresses. Cells are also able to differentiate into morphological forms better adapted to survive in hostile environment (Smith et al. 2003). A common autoinducer in gram negative bacteria, the N-acyl homoserine lactones (AHLs) is produced when cell density increases, and is critical in regulating genes important for dissemination and virulence in animal and plant pathogens (Molina 2003). AHLs are one of the most widespread and best understood families of signal molecules among Gram-negative bacteria. These molecules can vary greatly with the presence or absence of an acyl chain C3 substituent (oxoor hydroxy-) and length of the N-acyl side chain (four to 14 carbons) (Fray 2001). AHL regulated phenotypes are capable of facilitating interactions between the producing organism and the surrounding environment, including both pathogenic and symbiotic relationships with higher organisms (Manefield et al. 2001). AHLs are known to regulate the diverse enzymes and toxins produced by E. carotovora. CarI of E. carotovora produces 3-oxo-C6-HSL, which is responsible for the induction of the secreted plant cell wall-degrading exoenzymes and of the antibiotic carbapenem (Fray 2001). One method used to disrupt quorum sensing in E. carotovora is the introduction of the aiiA gene cloned from Bacillus sp. into transgenic tobacco and potato plants (Molina 2003). In a tobacco test system, E. carotovora carI mutants are completely avirulent. They are incapable of macerating plant tissue or multiplying in planta since they lack pectin lyase, pectate lyase, polygalacturonase, cellulase and

PAGE 29

19 protease (Fray 2001). Once the gene was expressed and AHL-lactonase were produced, the quorum sensing systems were paralyzed which resulted in increased plant disease resistance (Molina 2003). Another method involves constructing transgenic tobacco lines that express the E. carotovora AHL gene, expI. Ectopic production of bacterial AHL by the transgenic plants tricks the pathogen into prematurely secreting virulence factors, such as pectinolytic enzymes, when cell populations are insufficient for infection. This is thought to trigger host plant defenses resulting in the observed disease resistance. Since genetically modified crops are not accepted in many countries, a better strategy would be the application of microorganisms with natural ability to degrade AHLs (Molina 2003).

PAGE 30

CHAPTER 3 MATERIALS AND METHODS Whole and injured tomatoes were used. Tomato surfaces and shave-wounded blossom scars were inoculated with several bacterial cocktails. These included a Salmonella cocktail comprised of five rifampicin-resistant serovars, a Shigella cocktail comprised of two rifampicin-resistant species and a serovar of Erwinia carotovora resistant to rifampicin. There were several sets of inoculated tomatoes including an Erwinia and Salmonella cocktail combination, Erwinia and Shigella cocktail combination, Salmonella cocktail alone, Shigella cocktail alone, and Erwinia alone. There was also an uninoculated group of tomatoes to serve as the control. Recovery of the pathogens from tomato surfaces and shave-wounded blossom scars was monitored on Days 0, 1, 2, and 3. The tomatoes were maintained at three different temperature and relative humidity conditions to simulate various storage conditions. Selection of Temperature and Relative Humidity Conditions The Florida Automated Weather Network (FAWN) (University of Florida Institute of Food and Agricultural Sciences 2003) weather archives allowed for the selection of temperature and relative humidity settings to represent the Florida fall/winter tomato production season and ripening room conditions (Table 3-1). To determine the effects of humidity on the growth and survival of Erwinia, the temperature was kept at an optimum with varying humidity to represent standard ripening rooms and Florida fall/winter tomato production season. The chosen parameters were used to simulate an open-air packinghouse environment. 20

PAGE 31

21 Table 3-1. Temperature and relative humidity conditions selected to simulate a standard ripening room environment (90% RH, 20C) and optimum temperature for the growth of Erwinia at fall/winter conditions (60% RH, 27C) and a ripening room (90% RH, 27C) Simulated environment Relative humidity (%) Temperature (C) Standard tomato ripening room 90 20 Erwinia optimum temperature with tomato ripening room humidity 90 27 Erwinia optimum temperature with Florida fall/winter production season humidity 60 27 Acquisition and Maintenance of Salmonella Cultures Salmonella serovars were received from Dr. Linda J. Harris (University of California, Davis, Department of Food Science and Technology). My study used five Salmonella enteritidis serovars; Agona, Gaminara, Michigan, Montevideo, and Poona (Table 3-2). The serovars obtained were adapted to rifampicin (rif+) at the University of Florida using methods described by Lindeman and Suslow (1987). The five Salmonella serovars (rif+) were transferred to PROTECT Bacterial Preservers (Scientific Device Laboratories, Des Plaines, IL) and stored at -70C. The serovars were also transferred to NA (rif+) slants and stored at 4C. Rifampicin is an antibiotic capable of inhibiting protein synthesis of mammalian cells and it is freely soluble in methanol (Merck Index 2001). A 10,000 ppm (1%) stock solution of rifampicin was utilized throughout this study. The stock solution was prepared by dissolving 1.0 g of rifampicin (ICN Biomedical, Inc., Aurora, OH) dissolved in 100 mL of high performance liquid chromatography (HPLC) grade methanol (Fisher Scientific International, Fair Lawn, NJ). The stock solution was filter sterilized, protected from light, and stored at room temperature. The Salmonella serovars were adapted to 200 ppm rifampicin (rif+) initially and then lowered to 80 ppm rifampicin

PAGE 32

22 (rif+) during experimentation to decrease the amount of stress placed on the pathogen while still maintaining selectivity. Nutrient Agar (NA) (BD, Sparks, MD), supplemented with 80 g/mL rifampicin (rif+) antibiotic was used to recover Salmonella from inoculated surfaces. The rifampicin marker allowed differentiation of inoculated Salmonella from naturally present bacteria that may have been present on the samples (Burnett and Beuchat 2001; Lukasik et al. 2001). Table 3-2. Salmonella enteritidis serovars obtained from Dr. Linda J. Harris at the University of California, Davis: wild type serovars listed with source Serovar Designation Serovar Name Origin LJH517 Agona Alfalfa sprouts LJH518 Gaminara Orange juice LJH521 Michigan Cantaloupe LJH519 Montevideo Human isolate from tomato outbreak LJH630 Poona Human isolate from tomato outbreak Acquisition and Maintenance of Shigella Cultures Shigella flexneri (LJH607) was obtained from Dr. Linda Harris (University of California, Davis, Department of Food Science and Technology). Shigella sonnei (ATCC 9290) was purchased from the American Type Culture Collection (ATCC, Manassas, VA). The serovars obtained were adapted to rifampicin (rif+) at the University of Florida using methods described by Lindeman and Suslow (1987). The two Shigella serovars (rif+) were transferred to PROTECT Bacterial Preservers and stored at -70C. The Shigella species were also transferred to NA (rif+) slants and stored at 4C. The rifampicin was prepared using the same method as stated above. The Shigella serovars were adapted to 200 ppm rifampicin (rif+) and then reduced to 80 ppm during experimentation to minimize stress on the cells while maintaining selectivity. Nutrient Agar (NA) supplemented with 80g/mL rifampicin (rif+) antibiotic was used to recover Shigella from inoculated surfaces. The use of rifampicin resistant Shigella

PAGE 33

23 allowed for differentiation from background bacteria and only allow for growth of those pathogens adapted to rifampicin (rif+). Acquisition and Maintenance of Erwinia carotovora Culture Erwinia carotovora subsp. carotovora (SR38) was obtained from Dr. Jerry A. Bartz (University of Florida, Department of Plant Pathology). This strain of Erwinia carotovora (SR38) was obtained from a shipment that was rejected due to decay. This serovar was adapted rifampicin (rif+) at the University of Florida using methods described by Kaspar and Tamplin (1993). The serovar was then transferred to PROTECTTM Bacterial Preservers and stored at -70C. Erwinia carotovora was also transferred to NA (rif+) slants and stored at 4C. The Erwinia were adapted to a maximum concentration of 200 ppm (rif+) and then lowered to 80 ppm during experimentation to allow for growth while still being a selective medium. Nutrient Agar (NA) supplemented with 80g/mL rifampicin (rif+) antibiotic was used to recover Erwinia from inoculated surfaces. Rifampicin (rif+) was used to eliminate any background bacteria that may have been present on the samples. Growth Studies Growth studies were performed in triplicate for each pathogen and its respective serovars to determine the rate of growth for each. These studies were done to ensure that each cocktail contains the same quantity of each serovar (CFU/mL), preventing domination of one serovar over the others. The five Salmonella serovars (rif+), two Shigella serovars (rif+) and the Erwinia carotovora serovar (rif+) were revived off PROTECT Bacterial Preservers by aseptically transferring one bacterial preserver into 10 mL of Tryptic Soy Broth (TSB) (BD, Sparks, MD) supplemented with 80 l of rifampicin. The Salmonella and Shigella cultures were then incubated in a shaking

PAGE 34

24 incubator (Queue Systems, Asheville, NC) at 30 rotations per minute at 37C for 24 hours. The Erwinia culture was incubated on a desktop shaking incubator (Thermolyne RotoMix Type 50800, Dubuque, IA) at 100 rotations per minute at 25-27C for 24 hours. The cultures were successively transferred for two days in 10 mL of fresh TSB (rif+) to obtain uniform cell type (Beuchat et al. 2001) and incubated again at its respective temperature for 24 hours. Three 100 mL flasks of fresh TSB (rif+) were prepared. A loop of each serovar was used to inoculate each of the flasks. Every hour a cuvette reading of the culture (1 mL) was obtained and read in a spectrophotometer (Shimadzu Scientific Instruments, Inc., Model UV-1201, Japan) set at 600 Following three successive hourly cuvette readings, the growth curve was concluded due to the cells reaching the stationary phase. Additionally, every hour 1 mL of the inoculated TSB (rif+) was serially diluted (1:10) in 9 mL tubes of sterile 0.1% peptone water (BactoTM Peptone, Sparks, MD). The appropriate dilutions were plated out using the pour plate technique for the Salmonella and Shigella serovars and the spread plate technique for the Erwinia carotovora using NA (rif+). Plates for the cultures were allowed to set and inverted to be statically incubated at their respective temperatures for 24-48 hours. Colony forming units (CFU) were observed and counted. Preparation of Inoculum Three days prior to the beginning of each experiment, the pathogens were revived off of TSA (rif+) slants. Overnight transfers were performed using 10 mL tubes of TSB (rif+) each day when growth was visible. On the day of the experiment, cells were washed twice via centrifugation (4,000 x g, 10 minutes at 4C) using 0.1% peptone water. Equivalent aliquots of the five serovars of Salmonella were combined as a cocktail. The same procedure was used to make the cocktail for the two species of

PAGE 35

25 Shigella. The Erwinia carotovora, Shigella and Salmonella cultures had reached populations of approximately 1.0 x 108 CFU/mL. When making the combined cocktails containing either Salmonella or Shigella with Erwinia, half of the cocktail contained the plant pathogen while the other half consisted of the previously made cocktail of the Salmonella or Shigella. The inoculum was serially diluted using 9 mL tubes of 0.1% peptone water to confirm cell concentrations. The dilutions were plated in triplicate using the pour plate technique for Salmonella and Shigella and the spread plate technique for Erwinia carotovora using NA (rif+). Inoculation of Tomatoes Intact Tomatoes Unwashed, unwaxed mature green tomatoes, variety Florida 47, were supplied by Six Ls Packing, Inc. (Immokalee, FL) and DiMare (Tampa, FL) for all experimental studies. Tomatoes were classified as 6x7 (formerly medium) by the Florida Tomato Committee (Florida Tomato Committee 2005). When preparing for inoculation, tomatoes were placed aseptically on sterile fiberglass trays with the stem scars facing down. The tomatoes were sprayed with reagent alcohol and wiped with a Kimwipe (Kimberly-Clark, Neenah, WI) to remove any particulate matter from the surface of the tomato. Once clean and dry, the tomatoes were inoculated around the blossom scar area, not directly on the blossom scar, using a Repeater Plus pipette (Eppendorf AG, Germany). The tomato inoculation consisted of ten 10 l spots of the pathogen suspension yielding a total inoculation of 100 l per whole tomato. Immediately after inoculation, the intact tomato samples were allowed to dry completely at room temperature on the lab bench. Once dry, the samples needed for that day were taken and the rest of the tomatoes were placed in a Caron 6030 (Caron, Marietta, OH)

PAGE 36

26 environmental humidity chamber. The Caron humidity chamber was equipped with a Caron CRS 101 water supply system which utilized distilled water to humidify the chamber. The temperature and relative humidity inside the chamber was continuously controlled and displayed by a Whatlow Series 96 temperature and relative humidity controller (Whatlow, Winona, MN). Shave-Wounded Tomatoes Unwashed, unwaxed mature green tomatoes were prepared for inoculation in a similar manner as mentioned above. A sharpened knife was sterilized by spraying with reagent alcohol and putting it through a flame. Once the knife was cool to the touch, three shave wounds were made by slicing a 1-2 mm section around the blossom end of the tomato. Between each shaving, the knife was sterilized and a total of three shave wounds were made on each tomato. The bacteria were inoculated directly onto the shave wounds using ten 10 l spots yielding a total of 100 l of suspension. The samples needed for that day were taken immediately and the rest were placed in a Caron 6030 environmental humidity chamber with the same functions as mentioned previously. Pathogen Recovery off Tomato Surfaces Intact Tomatoes Tomatoes were sampled from the environmental humidity chamber on Days 0, 1, 2 and 3 and recovery studies were performed. Each days sample consisted of three replicates. On Day 0, once the inoculum was dry and before being put in the environmental humidity chamber, the samples for that day were immediately placed in sterile Stomacher (Seward, West Sussex, UK) bags containing 100 mL of sterile 0.1% peptone water. For sampling on Days 1, 2 and 3, the tomatoes were aseptically removed from the environmental humidity chamber and individually placed into sterile

PAGE 37

27 Stomacher bags containing 100 mL of sterile 0.1% peptone water. Tomato samples were rubbed and shaken for one minute (Burnett and Beuchat 2001; Harris et al. 2001; Zhuang et al. 1995) with rubbing action concentrated around the inoculated blossom scar area of the tomatoes to recover any attached bacteria. The sample from each Stomacher bag was then serially (1:10) diluted using sterile 0.1% peptone water dilution tubes. The serial dilutions were pour-plated for the Salmonella and Shigella cocktails and spread plated for Erwinia, Salmonella-Erwinia and Shigella-Erwinia combinations using NA (rif+). A negative control for the NA (rif+) was poured in duplicate to ensure the media was not contaminated. The plates were inverted and statically incubated at their respective temperatures for 48 hours. Shave-Wounded Tomatoes Tomato shave wounds were sampled on Days 0, 1, 2 and 3 and recovery studies were performed. Each days sample consisted of three replicates. On Day 0, immediately after inoculating the shave wound, the knife was flame sterilized with reagent alcohol and used to remove a slice 1-2 mm in thickness at the site of inoculation. Once these initial samples were taken, the shave wound tomatoes were placed in the environmental humidity chamber until future sampling. Each day, samples were cut from the shave wounded tomato and immediately put in sterile Stomacher bags containing 100 mL of sterile 0.1% peptone water. Tomato samples were stomached (AES Laboratoire, Comourg, France) for one minute. The samples from each Stomacher bag were then serially (1:10) diluted using sterile 0.1% peptone water dilution tubes. The appropriate serial dilutions for the Salmonella and Shigella cocktail were pour plated, while the Erwinia, Salmonella-Erwinia and Shigella-Erwinia combinations were spread

PAGE 38

28 plated using NA (rif+). A negative control of NA (rif+) was plated in duplicate to ensure the media was not contaminated. Uninoculated whole tomatoes and shave wounded tomatoes were used as control samples throughout the length of the studies. The intact tomato control samples were placed in sterile Stomacher bags of 0.1% peptone water, rubbed and shaken and serially (1:10) diluted as mentioned previously. The shave wounded control tomatoes were sampled using a flame sterilized knife and stomached for one minute in 0.1% peptone water. The samples were then serially (1:10) diluted and inverted to be statically incubated at room temperature (25-27C) for 48 hours. Shave-Wounded Tomatoes with Combined Pathogens Shave wounded tomatoes were inoculated with the pathogen combinations of Salmonella-Erwinia or Shigella-Erwinia. Samples were obtained immediately after inoculation on Day 0, and then approximately 24 hours later on each Days 1, 2, and 3. All tomato samples were done in triplicate to ensure repeatability. Samples were obtained in the same manner mentioned in the previous section. Once stomached, samples from the tomato slices were serially diluted appropriately and plated on both NA (rif+) and Salmonella-Shigella (SS) Media. It was previously determined that Erwinia was not capable of growth on SS Media. In determining the viable bacterial counts, the NA (rif+) plates represented the total amount of bacteria recovered (i.e., Salmonella or Shigella and Erwinia). The SS Media represented either Salmonella, detectable by black colonies, or Shigella, apparent as red/orange colonies. To obtain the counts for Erwinia, the NA (rif+) plate counts (total population) were subtracted from the SS Media plate counts (Salmonella or Shigella).

PAGE 39

29 Statistical Analysis All results from the longevity studies were average log10 counts (log10 CFU/mL) of recovered Salmonella, Shigella and Erwinia. Statistical analyses were performed with the Statistical Analysis System (SAS Institute, Cary, NC). The GLM (general linear model) procedure in SAS was used to evaluate the significance between environmental conditions, time (days), the condition of the tomato surface and combinations thereof for each pathogen utilized. Multiple comparisons were performed using the LS (least squares) Mean method. Results with P<0.05 were considered significant in these longevity studies. In determining significance of the shave-wounded bacteria combination tomatoes, two tailed t-tests with P<0.05 was considered significant.

PAGE 40

CHAPTER 4 RESULTS Recovery studies were performed to determine the survival of Salmonella, Shigella and Erwinia on intact and shave-wounded tomato surfaces. Inoculated fruit were subjected to three different temperature/relative humidity conditions for three days. The three environments consisted of standard ripening room conditions (90% RH, 20C), Florida fall/winter production line conditions (60% RH, 27C), and optimum conditions for the survival and growth of Erwinia carotovora subspecies carotovora (90% RH, 27C). Inoculated fruit were sampled daily to determine the survival/growth of the bacteria. The recovery rate of these bacteria was used to assess the optimum environmental conditions and tomato surface condition that would allow for survival and growth. Additionally, the combination recovery studies were used to determine if there was a correlation between the presence of Erwinia and growth or survival of the foodborne bacteria Salmonella or Shigella over time. Each inoculation group was sampled in groups of three to five at specific intervals; Day 0, 1, 2, and 3. Intact tomato samples consisted of the whole tomato, while the shave-wounded tomato samples contained a 1-2 mm shaved off portion that had been inoculated. The recovered bacteria from each of the three replicates were counted and averaged. Next, the data was compiled into graphs depicting the relationship between log10 CFU/mL of the bacterial survivors and time (days) for either intact or shave-wounded tomatoes in each simulated environment. 30

PAGE 41

31 Growth Levels of Salmonella, Shigella and Erwinia carotovora Growth studies were performed in triplicate, for each of the five rifampicin-resistant Salmonella serovars (Figure 4-1), two rifampicin-resistant Shigella species (Figure 4-2) and rifampicin-resistant Erwinia carotovora (Figure 4-3). 024681012012345678910Time (hours)Average Log10 CFU/ml S. Agona S. Michigan S. Gaminara S. Montevideo S. Poona Figure 4-1. Average log10 CFU/mL growth curves of Salmonella serovars over a 10-hour incubation. 0123456789100123456789101112Time (hours)Average Log10 CFU/ml S. sonnei S. flexneri Figure 4-2. Average log10 CFU/mL growth curves of Shigella species over a 12-hour incubation.

PAGE 42

32 0123456789100123456789101112131415161718192021Time (hours)Average Log10 CFU/ml E. carotovora Figure 4-3. Average log10 CFU/mL growth curves of Erwinia carotovora over a 21-hour incubation. Results from these preliminary studies ensured that serovar growth rates were equivalent to one another and a consistent inoculum suspension could be made (Figure 4-4). The cell concentration of each bacterial cocktail was estimated before beginning any experiments. To determine the suspension concentration the cocktail was pour plated for Salmonella and Shigella and spread plated for Erwinia using the appropriate dilutions in triplicate on NA (rif+). 012345678910S. AgonaS. Gaminara S. MichiganS. MontevideoS. PoonaShigella flexneriShigella sonneiErwinia carotovoraAverage Log CFU/ml Figure 4-4. Maximum average log10 CFU/mL counts of Salmonella serovars, Shigella species and Erwinia (rif+) after specific incubation times.

PAGE 43

33 Preliminary Recovery Studies: Intact Tomato Surfaces Before beginning the longevity studies on intact and shave-wounded tomatoes, tomatoes were inoculated with the five-strain Salmonella cocktail, two-strain Shigella cocktail or Erwinia carotovora. The intact tomatoes were sampled once the inoculum had dried on the surface. Each tomato underwent the shake-rub method and was plated using the pour plate method for the Salmonella and Shigella and the spread plate method for Erwinia. All bacteria were recovered from the tomato surface with no more than a 1.5 log10 CFU/mL reduction from the initial inoculation concentration (Table 4-1). Intact tomatoes that were not inoculated served as the control and were used to ensure that the NA (rif+) media was sufficient in eliminating background bacteria. Table 4-1. The log10 CFU/mL reduction of individual bacterias off intact tomato surfaces during preliminary studies Log CFU/mL SD Salmonella Shigella Erwinia Initial inoculation 5.57 0.02 5.08 0.11 4.97 0.17 Recovery post drying 4.47 0.15 3.65 0.35 4.94 0.17 Total reduction 1.10 0.15 1.43 0.35 0.03 0.17 Recovery of Bacteria from Tomato Surfaces Mature green tomatoes (Florida 47) were inoculated with the individual bacterial suspensions or combinations on both intact and shave-wounded tomatoes. Samples were stored for three days in an environmental humidity chamber to simulate specific parameters. Tomatoes not inoculated served as the controls and were sampled each day to ensure the rifampicin was efficient in eliminating background microflora off the surface or shave-wound of the tomato samples. All control tomatoes were found to be negative.

PAGE 44

34 Recovery of Bacteria from Intact Tomato Surfaces Salmonella Recovery off Intact Tomato Surfaces Intact tomatoes were subjected to the three simulated environments. The initial Salmonella inoculum concentration applied to all intact tomato surfaces was approximately 8.09 log10 CFU/mL. All intact tomatoes inoculated with the Salmonella cocktail subjected to the three simulated environments showed an overall decrease in log10 CFU/mL values from Day 0 to 3 (Figure 4-5). The smallest log10 CFU/mL reduction was shown to occur with the 90% RH, 27C conditions with a total decrease of 0.64 log10 CFU/mL. The reduction of Salmonella from tomatoes subjected to the 90% RH, 20C conditions was 4.43 log10 CFU/mL, while the reduction of Salmonella off of the tomatoes subjected to the 60% RH, 27C conditions was 3.59 log10 CFU/mL. From LS Means analysis, it was shown that the 60% RH, 27C had significantly lower recovery than both environments kept at 90% RH (P<0.05). There were no significant differences observed between the intact tomatoes inoculated with Salmonella that were subjected to 90% humidity regardless of temperature (20 or 27) (P>0.05). For all intact Salmonella studies, there were no significant differences observed between environmental conditions on Day 0. The recovered Salmonella was significantly less under 60% RH, 27C environmental conditions than the 90% RH, 27C and 90% RH, 20C on Day 1. All three environmental conditions were significantly different from each other on Days 2 and 3 (P<0.05). It can be observed that Salmonella is capable of surviving on intact tomato surfaces over three days with higher humidity and temperature (90% RH, 27C) allowing for the most recovery.

PAGE 45

35 01234560123Time (days)Log CFU/ml Survivor s 60% RH, 27C 90% RH, 27C 90% RH, 20C Figure 4-5. Salmonella recovery (log10 CFU/mL) from intact tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days. Shigella Recovery from Intact Tomato Surfaces Intact tomatoes were subjected to the same three environmental conditions as explained above. The initial inoculation concentration applied to the intact tomatoes was approximately 8.04 log10 CFU/mL. All intact, inoculated tomatoes displayed an average log10 CFU/mL decrease over the three day study (Figure 4-6). The greatest log10 CFU/mL decrease occurred under the standard ripening room conditions (90% RH, 20C) with a 4.39 log10 CFU/mL reduction. The recovery under fall/winter conditions (60% RH, 27C) resulted in a 3.69 log10 CFU/mL reduction, while the optimum conditions for Erwinia (90% RH, 27C) resulted in a 1.36 log10 CFU/mL reduction. There was a significant increase in recovered Shigella on Day 1 held at the optimum conditions for Erwinia (90% RH, 27C). There was also a slight increase in log10 CFU/mL shown on Day 2 of the fall/winter conditions (60% RH, 27C). LS Means analysis showed that there were no significant differences between the fall/winter season conditions (60% RH, 27C) and the standard ripening room conditions (90% RH, 20C).

PAGE 46

36 The recovery of Shigella under optimum conditions for Erwinia (90% RH, 27C) was significantly greater than the other two treatments. There were no significant differences in recovered Shigella found between Days 1, 2 and 3 under fall/winter and standard ripening room conditions. The least recovery occurred under fall/winter season conditions (60% RH, 20C) and the standard ripening room conditions (90% RH, 20C). Shigella was most recoverable on intact tomato surfaces for three days under the optimum conditions for the growth of Erwinia (90% RH, 27C). 012345600.511.522.53Time (days)Log CFU/ml Survivors 60% RH, 27C 90% RH, 27C 90% RH, 20C Figure 4-6. Shigella recovery (log10 CFU/mL) from intact tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days. Erwinia Recovery from Intact Tomato Surfaces Intact tomatoes inoculated with an initial concentration of approximately 8.15 log10 CFU/mL of Erwinia carotovora were subjected to the same three environmental conditions as explained previously. Erwinia subjected to all three environmental conditions showed an average log10 CFU/mL decrease over the three day study (Figure 4-7). The greatest decrease occurred under standard ripening room conditions (90% RH, 20C) with a reduction of 5.71 log10 CFU/mL. The recovery under fall/winter conditions

PAGE 47

37 (60% RH, 27C) resulted in a 4.58 log10 CFU/mL reduction, while the optimum conditions for Erwinia resulted in a 1.36 log10 CFU/mL decrease. There was a slight increase in recovery observed on Day 2 under optimum conditions for Erwinia (90% RH, 27C). There were no significant differences observed between the standard ripening room conditions (90% RH, 20C) and the optimum conditions for Erwinia carotovora (90% RH, 27C) on Day 0 according to LS Means analysis. There were no significant differences observed between Day 1 and Day 2 for the intact tomatoes inoculated with Erwinia. Erwinia was able to survive for the span of the three day study under its optimum conditions for growth (90% RH, 27C) on intact tomato surfaces. The fall/winter season conditions (60% RH, 27C) had no recovery of Erwinia after Day 0. The lower temperature (20C) and relative humidity (60%) was an important factor in the recoverability of the Erwinia off of intact tomato surfaces. 012345670123Time (days)Log CFU/ml Survivors 60% RH, 27C 90% RH, 27C 90% RH, 20C Figure 4-7. Erwinia carotovora recovery (log10 CFU/mL) from intact tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days.

PAGE 48

38 Recovery of Bacteria from Shave-Wounded Tomato Surfaces Salmonella Recovery from Shave-Wounded Tomato Surfaces Shave-wounded tomatoes were subjected to three environmental conditions consisting of fall/winter tomato season (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C), and standard ripening room conditions (90% RH, 20C). The inoculum was serially diluted to allow for growth on the wounded tomato surface. Each shave-wound was inoculated with approximately 5.46 log10 CFU/mL of the five-strain Salmonella cocktail. All shave-wounded tomato surfaces showed an average log10 CFU/mL increase during the three day experiments (Figure 4-8). The greatest log10 CFU/mL increase occurred under standard ripening room conditions (90% RH, 20C) with an increase of 4.00 log10 CFU/mL. The optimum conditions for Erwinia (90% RH, 27C) resulted in a 3.35 log10 CFU/mL increase, while the fall/winter conditions (60% RH, 27C) showed a 1.26 log10 CFU/mL increase. 012345670123Time (days)Log CFU/ml Survivors 60% RH, 27C 90% RH, 27C 90% RH, 20C Figure 4-8. Salmonella recovery (log10 CFU/mL) from shave-wounded tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days. Results from LS Means analysis showed that significantly fewer Salmonella were recovered at 60% RH, 27C when compared to both 90% RH, 27C and 90% RH, 20C

PAGE 49

39 (P<0.05). Under higher humidity conditions (90% RH) there were no significant differences observed on Day 2 and 3, showing that humidity and not temperature is an important factor in the survival and growth of Salmonella on wounded tomato surfaces. Shigella Recovery from Shave-Wounded Tomato Surfaces The initial Shigella inoculum was serially diluted to allow for growth on the wounded tomato surface. Shave-wounded tomatoes were inoculated with an initial concentration of approximately 5.04 log10 CFU/mL and subjected to the same three environmental conditions as explained previously. There was no recovery observed from any of the samples obtained under fall/winter conditions (60% RH, 27C), while the standard ripening room (90% RH, 20C) and optimum conditions for Erwinia (90% RH, 27C) resulted in a slight log10 CFU/mL increase (Figure 4-9). Shigella subjected to standard ripening room conditions resulted in a 2.35 log10 CFU/mL increase and the optimum conditions for Erwinia resulted in a 1.28 log10 CFU/mL increase over the span of three days. On Day 1 of the Erwinia optimized conditions (90% RH, 27C), there was a relatively large increase in log10 CFU/mL for Shigella before decreasing slightly on Day 2 and remaining constant for the remainder of the study. There were no significant differences observed between the standard ripening room conditions (90% RH, 20C) and the optimum conditions for Erwinia (90% RH, 27C) over three days. The recovery of Shigella under fall/winter conditions was significantly less than the environmental conditions held at higher humidity (90%). Under the fall/winter season conditions, a possible explanation for the lack of recovered cells may be that the inoculum did not sufficiently attach to the wounded tomato surface before being placed in the environmental humidity chamber. Once inside the humidity chamber,

PAGE 50

40 the wounded tomato surfaces showed visible desiccation within one day preventing the survival of any Shigella that may have been present. 0123450123Time (days)Log CFU/ml Survivors 60% RH, 27C 90% RH, 27C 90% RH, 20C Figure 4-9. Shigella recovery (log10 CFU/mL) from shave-wounded tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days. Erwinia Recovery from Shave-Wounded Tomato Surfaces Shave-wounded tomatoes were kept in three separate environmental conditions for three days as explained previously. The initial concentration of Erwinia carotovora was serially diluted and applied to all shave-wounded tomatoes at a concentration of approximately 5.64 log10 CFU/mL. The largest increase of Erwinia carotovora recovered occurred under optimum conditions (90% RH, 27C) resulting in a 6.59 log10 CFU/mL increase (Figure 4-10). The fall/winter conditions resulted in a 2.59 log10 CFU/mL increase, while the standard ripening room conditions resulted in a 5.93 log10 CFU/mL increase. There were no significant differences observed between the standard ripening room conditions (90% RH, 20C) and the optimum conditions for Erwinia (90% RH, 27C) according to LS Means analysis. The recovery of Erwinia under fall/winter conditions (60% RH, 27C) was significantly less than the environments held at higher humiditys

PAGE 51

41 (90% RH, 27C and 90% RH, 20C). The lower humidity (60%) used for fall/winter conditions allowed for some growth of Erwinia but did not show the exponential growth displayed by those environments kept at a higher humidity (90%). A high relative humidity is important in allowing for survival and growth of Erwinia carotovora. 0123456789100123Time (days)Log CFU/ml Survivors 60% RH, 27C 90% RH, 27C 90% RH, 20C Figure 4-10. Erwinia carotovora recovery (log10 CFU/mL) from shave-wounded tomato surfaces in fall/winter (60% RH, 27C), optimum conditions for Erwinia (90% RH, 27C) and ripening room parameters (90% RH, 20C) over three days. Recovery of Combined Bacteria from Shave-Wounded Tomato Surfaces Recovery of Salmonella and Erwinia in Fall/Winter Season Conditions Shave-wounded tomatoes were kept in the environmental humidity chamber for the length of the study at the fall/winter season conditions (60% RH, 27C). The initial concentration of the Salmonella-Erwinia combination was serially diluted and applied to the shave-wounds at a concentration of approximately 5.35 log10 CFU/mL. The concentration of the inoculation was reduced to allow for growth. Both the Salmonella and Erwinia counts increased throughout the three-day study resulting in a 3.99 log10 CFU/mL increase of Erwinia and a 2.72 log10 CFU/mL increase of Salmonella (Table 4-2).

PAGE 52

42 According to the two tailed t-test, there were significant increases observed for Salmonella between Day 1 and 2 of the study. The combined bacteria under fall/winter season conditions allowed for a greater amount of growth of both microorganisms than when individually inoculated onto the wounded tomato surface as seen at the Day 2 sampling period and beyond. These effects show that the bacteria may be creating more favorable environments for each other by Erwinia increasing the moisture content, releasing nutrients and Salmonella possibly providing a protective covering to the lower humidity (60%) when combined. Table 4-2. Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in fall/winter (60% RH, 27C) conditions over three days. Log10 CFU/mL SD Salmonella individually Salmonella recovered in combination with Erwinia Erwinia individually Erwinia recovered in combination with Salmonella Day 0 0.70 1.09a 1.85 0.52a 0.26 0.58a 2.13 0.66b Day 1 0.92 0.95a 1.59 1.34a 1.98 1.14a 2.67 1.14a Day 2 1.42 1.37a 4.22 0.97b 1.91 2.01a 6.00 0.54b Day 3 2.00 1.68a 4.56 0.80b 2.85 2.67a 6.11 0.37b Note: Letters (a,b) following Log10 CFU/mL SD among bacteria (i.e., Salmonella vs. Salmonella combined) in rows are significantly different. Recovery of Salmonella and Erwinia in Optimum Conditions for Erwinia Shave-wounded tomatoes were kept in the environmental humidity chamber for the length of the study at the conditions considered optimal for Erwinia carotovora (90% RH, 27C). The initial concentration of the applied inoculum of Salmonella-Erwinia was approximately 4.87 log10 CFU/mL. This inoculum concentration was obtained by serially diluting the stock culture, which would allow for sufficient growth on the wounded tomato surface. The Salmonella and Erwinia grew substantially over the period

PAGE 53

43 of three days resulting in a 4.94 log10 CFU/mL increase of Erwinia and a 4.81 log10 CFU/mL increase of Salmonella (Table 4-3). There were no significant differences observed in growth Salmonella when comparing the individual counts and those recovered in combination with Erwinia on Days 1,2 or 3. In contrast, the Erwinia grew to a significantly lesser amount when combined with the Salmonella. This could be due to competition between the microorganisms Salmonella and Erwinia. Table 4-3. Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in optimized conditions for Erwinia carotovora (90% RH, 27C) over three days. Log10 CFU/mL SD Salmonella individually Salmonella recovered in combination with Erwinia Erwinia individually Erwinia recovered in combination with Salmonella Day 0 1.98 0.23a 1.45 0.38b 2.21 0.08a 1.65 0.46a Day 1 4.27 0.47a 5.18 0.98a 5.92 0.32a 5.61 0.99a Day 2 5.54 1.39a 5.20 1.66a 7.93 0.17a 6.34 1.10b Day 3 5.32 0.23a 6.24 1.21a 8.80 0.17a 6.59 0.45b Note: Letters (a,b) following Log10 CFU/mL SD among bacteria (i.e., Salmonella vs. Salmonella combined) in rows are significantly different. Recovery of Salmonella and Erwinia in Standard Ripening Room Conditions Shave-wounded tomatoes were kept in an environmental humidity chamber for three days set to the conditions of a standard ripening room (90% RH, 20C). The stock culture was serially diluted before being inoculated onto the shave-wounded tomatoes to allow for growth. The initial concentration of the Salmonella-Erwinia combination inoculum was 5.14 log10 CFU/mL. The Salmonella and Erwinia grew daily with the biggest increase seen between Day 0 and Day 1 (Table 4-4). The Erwinia had a total increase of 6.26 log10 CFU/mL while the Salmonella had a 4.18 log10 CFU/mL increase over three days.

PAGE 54

44 Growth of Salmonella combined with Erwinia was not significantly different from Salmonella individually inoculated onto shave-wounded tomatoes except on Day 3; growth was significantly less. The same pattern of significance was seen for Erwinia. Both Salmonella and Erwinia were recovered in lower numbers when combined under standard ripening room conditions (90% RH, 20C). Although growth was observed when combined, the microorganisms may be competing with each other for nutrients and available space resulting in less growth. Table 4-4. Recovered Salmonella and Erwinia from shave-wounded tomato surfaces in standard ripening room conditions (90% RH, 20C) over three days Log10 CFU/mL SD Salmonella individually Salmonella recovered in combination with Erwinia Erwinia individually Erwinia recovered in combination with Salmonella Day 0 2.08 0.17a 0.92 0.85a 2.12 0.18a 1.89 1.13a Day 1 3.68 1.03a 4.83 0.07a 6.04 0.11a 5.44 1.07a Day 2 5.53 0.78a 4.55 0.06a 7.05 0.32a 6.35 0.41a Day 3 6.03 0.48a 5.08 0.33b 8.04 0.10a 7.35 0.04b Note: Letters (a,b) following Log10 CFU/mL SD among bacteria (i.e., Salmonella vs. Salmonella combined) in rows are significantly different. Recovery of Shigella and Erwinia in Fall/Winter Season Conditions Shave-wounded tomatoes were placed in an environmental humidity chamber set to simulate Florida fall/winter tomato season conditions (60% RH, 27C) for three days. The concentration of the Shigella-Erwinia combination applied to the shave-wounded tomatoes was 5.00 log10 CFU/mL. On Day 1 and Day 2 of the study, there was no Shigella recovered (Table 4-5). Unexpectedly on Day 3, there was a considerable amount of Shigella recovered from the shave-wounded tomatoes. This was most probably due to the increase in moisture available after the Erwinia carotovora had time to break down tomato cell walls and release fluids. There was an increase seen in the

PAGE 55

45 Erwinia throughout the study resulting in a 4.27 log10 CFU/mL increase, while the Shigella had a total increase of 3.92 log10 CFU/mL. There were no significant differences between Day 1 and Day 2 for Shigella since none was recoverable on standard plating methods. Over three days, both Shigella and Erwinia displayed significant increases (P<0.05). Erwinia displayed greater growth when inoculated in combination with Shigella on the wounded tomato surface. Shigella was able to survive in combination with Erwinia under the lower relative humidity conditions unlike when inoculated alone on the wounded tomato surfaces. The Erwinia was able to provide greater moisture availability by breaking down the plant tissues of the tomato and releasing fluids, allowing the Shigella to survive and then grow when conditions were most favorable. The Shigella may have provided a protective barrier for the Erwinia under the low humidity conditions (60%). Under these conditions, there is a synergistic effect observed between Shigella and Erwinia in creating an optimum environment for themselves to allow for survival and growth. Table 4-5. Recovered Shigella and Erwinia from shave-wounded tomato surfaces in fall/winter season conditions (60% RH, 27C) over three days. ND: Non-detectable by lowest dilution plating methods. Log10 CFU/mL SD Shigella individually Shigella recovered in combination with Erwinia Erwinia individually Erwinia recovered in combination with Shigella Day 0 NDa 1.19 0.22b 0.26 0.58a 1.93 0.22b Day 1 NDa NDa 1.98 1.14a 1.96 0.74a Day 2 NDa NDa 1.91 2.01a 2.53 2.04a Day 3 NDa 5.08 0.75b 2.85 2.67a 6.20 0.22b Note: Letters (a,b) following Log10 CFU/mL SD among bacteria (i.e., Shigella vs. Shigella combined) in rows are significantly different.

PAGE 56

46 Recovery of Shigella and Erwinia under Optimum Conditions for Erwinia Shave-wounded tomatoes were placed in an environmental humidity chamber to simulate the optimized conditions for growth of Erwinia carotovora (90% RH, 27C). The stock culture was serially diluted before inoculating the shave-wounded tomatoes to allow for growth. The Shigella-Erwinia combination inoculum had an initial concentration of 4.87 log10 CFU/mL. There was a steady increase in growth seen throughout the study for both the Shigella and Erwinia (Table 4-6). By Day 3, the Shigella had an increase of 4.84 log10 CFU/mL while the Erwinia had an increase of 6.55 log10 CFU/mL. Table 4-6. Recovered Shigella and Erwinia from shave-wounded tomato surfaces under optimum conditions for Erwinia carotovora (90% RH, 27C) over three days. Log10 CFU/mL SD Shigella individually Shigella recovered in combination with Erwinia Erwinia individually Erwinia recovered in combination with Shigella Day 0 1.69 0.29a 0.69 0.58b 2.21 0.08a 1.00 0.89b Day 1 3.49 0.73a 0.92 0.27b 5.92 0.32a 4.02 0.58b Day 2 2.99 1.32a 3.22 1.63a 7.93 0.17a 6.19 1.06b Day 3 2.97 1.49a 5.51 0.39b 8.80 0.17a 7.55 0.55b Note: Letters (a,b) following Log10 CFU/mL SD among bacteria (i.e., Shigella vs. Shigella combined) in rows are significantly different. Shigella exhibited significant growth (P<0.05) over time in combination with Erwinia. Growth of Erwinia in combination with Shigella was significantly less in combination with Shigella. Under these environmental conditions, the Shigella grew to a much higher population when inoculated onto the wounded tomato surface in combination with Erwinia. There was significantly less growth of Erwinia observed on Day 3 when combined with Shigella. These results show that the Erwinia is allowing for more favorable conditions for the Shigella survival and growth by increasing moisture

PAGE 57

47 availability and possibly releasing nutrients previously not available before tissue maceration. Recovery of Shigella and Erwinia in Standard Ripening Room Conditions Shave-wounded tomatoes were subjected to ripening room conditions (90% RH, 20C) for a span of three days and inoculated with a Shigella-Erwinia combination inoculum with an initial concentration of 4.88 log10 CFU/mL. There were significant increases (P<0.05) for both Erwinia and Shigella over the span of the three day study (Table 4-7). The total increase of Shigella was 3.35 log10 CFU/mL while the Erwinia increased 6.39 log10 CFU/mL. Under standard ripening room conditions, Shigella and Erwinia recovered in combination was significantly less than when recovered individually. Since environmental conditions were not considerably stressful for either organism, it may be possible that the microorganisms were recovered at lesser concentrations when inoculated together due to competition with limited surface area for growth. Table 4-7. Recovered Shigella and Erwinia from shave-wounded tomato surfaces under standard ripening room conditions (90% RH, 20C) over three days. Log10 CFU/mL SD Shigella individually Shigella recovered in combination with Erwinia Erwinia individually Erwinia recovered in combination with Shigella Day 0 1.63 0.33a 0.35 0.58b 2.12 0.18a 1.82 1.00a Day 1 2.62 0.03a 2.64 0.19a 6.04 0.11a 2.91 0.12b Day 2 3.10 0.17a 2.94 0.13a 7.05 0.32a 6.52 0.48a Day 3 3.98 0.70a 3.68 0.21a 8.04 0.10a 7.52 0.18b Note: Letters (a,b) following Log10 CFU/mL SD among bacteria (i.e., Shigella vs. Shigella combined) in rows are significantly different.

PAGE 58

CHAPTER 5 DISCUSSION Consumption of fresh fruits and vegetables has greatly increased over the past 10 years (Lucier et al. 2001). The produce industry is continually confronted with the threat of microbial food safety hazards. Many steps are taken during the harvesting, processing and distribution of fresh produce, with the possibility of pathogenic contamination increasing significantly along the way. Environmental factors such as temperature and relative humidity have a large impact on the quality of fruits and vegetables along with the survival capacity of potential bacteria (Allen 2003). In recent years, multiple foodborne illnesses of Salmonella and Shigella have been associated with the consumption of contaminated produce, specifically tomatoes (Beuchat 1995; Brackett 1999). This study utilized a five-serovar rifampicin-resistant Salmonella cocktail, a two-species rifampicin-resistant Shigella cocktail and a rifampicin-resistant strain of Erwinia carotovora. These cocktails were applied individually and in combination to intact and shave wounded tomato surfaces. The tomatoes were subjected to different temperature and relative humidity combinations that simulated the conditions found in a standard ripening room (90% RH, 20C), fall/winter production conditions (60% RH, 27C) and optimum conditions for the growth of Erwinia (90% RH, 27C). Recovery of the bacteria from the intact tomatoes was performed by placing them into 100 mL of 0.1% peptone water and applying a rub-shake method (Burnett and Beuchat 2001), while the 48

PAGE 59

49 bacteria recovered from the shave wounded tomatoes were obtained by placing the sliced off wounded sections into 100 mL of 0.1% peptone water and stomaching for one minute. It has been recommended that a Salmonella cocktail should contain a minimum of five strains at approximately equal populations (CFSAN-FDA 2001). The five Salmonella enteritidis serovars selected for this study were S. Agona, S. Gaminara, S. Michigan, S. Montevideo and S. Poona. The serovars S. Agona, Gaminara and Michigan were isolated from fresh produce or their products (orange juice). S. Montevideo and Poona serovars were human isolates associated with fresh produce outbreaks. The two Shigella species selected for this study were S. flexneri and S. sonnei since they are the species most responsible for foodborne outbreaks. Erwinia carotovora subspecies carotovora was selected for this study because it is know to cause soft rot in many types of produce. All microorganisms were adapted to 200 g/mL of rifampicin. Rifampicin was selected because it is a stable marker and is effective in isolating bacteria from inoculated fruits that have natural background microflora. Growth characteristics for all microorganisms were evaluated by conducting growth studies. For the Salmonella serovars, S. Agona was observed to have the highest population at the end the 10-hour incubation period (37C), but there was less than a 0.6 log10 CFU/mL difference between the population of S. Agona and the serovar with the lowest growth level, S. Gaminara (Figure 4-1). All five serovars achieved counts of at least 1.0 x 108 CFU/mL after 10 hours of incubation. For the Shigella species, S. flexneri grew to the greatest population at the end of the 12-hour incubation period (37C), with less than a 0.5 log10 CFU/mL difference between the S. flexneri and S. sonnei growth (Figure 4-2). Both species achieved counts of at least 1.0 x 108 CFU/mL after the 12

PAGE 60

50 hour incubation. Erwinia carotovora grew to a population of at least 1.0 x 108 CFU/mL after the 21-hour incubation (Figure 4-3). These results showed that an acceptable inoculum could be prepared from each of the bacteria. Prior to each study, appropriate serial dilutions of the inoculum were pour-plated to determine the viable population of Salmonella and Shigella while the Erwinia was spread plated. These counts for each prepared inoculum for all experiments conducted also showed little variation between one another. All bacterial suspensions were determined to contain viable populations at 1.0 x 108 CFU/mL (Figure 4-4). Recovery of Bacteria from Intact Tomato Surfaces Intact tomatoes were subjected to three simulated environments including ripening room conditions (90% RH, 20C), fall/winter tomato production season parameters (60% RH, 27C) and optimum conditions for the growth of Erwinia carotovora (90% RH, 27C). Intact tomatoes have a firm surface that can withstand moderate rubbing and agitation so bacterial recovery from the tomato surface was performed by using a rub-shake method. This rub-shake method was chosen because it is the most effective technique for removing microorganisms from the surfaces of whole fruits and vegetables (FDA 2001). Intact, whole, unblemished tomatoes were specifically chosen for these inoculation studies. Spot inoculation of the cocktails was utilized because it allows for a known number of cells to be administered to the produce. Results from this study were similar to the findings of Guo et al. (2001) in that Salmonella populations decreased over time on intact tomato surfaces in all simulated environmental conditions (Figure 4-5). Salmonella was least recovered under standard ripening room conditions (90% RH, 20C), while recovery under fall/winter production conditions (60% RH, 27C) was not significantly greater (Figure 4-5). Between Day 0

PAGE 61

51 and Day 1, the largest log10 CFU/mL decrease of 2.53 log10 CFU/mL was seen under fall/winter production conditions (60% RH, 27C). Whereas, the most recovered Salmonella from intact tomato surfaces was seen under optimum conditions for Erwinia carotovora with only a 0.64 log10 CFU/mL decrease over three days. The higher humidity and temperature settings for the optimum conditions for growth of Erwinia allowed for the greatest recovery of the Salmonella cocktail. Lower humidity, such as what was used for the fall/winter production season conditions and the lower temperature used for the standard ripening room conditions limited the survival of the Salmonella cocktail on the intact tomato surface. Similar to the intact tomato studies inoculated with the Salmonella cocktail, there was an overall decrease of recovered Shigella over time (Figure 4-6). Both the standard ripening room conditions (90% RH, 20C) and the fall/winter production conditions (60% RH, 27C) produced significant decreases in recovered cells from Day 0 to Day 1. There was an unexpected increase of Shigella recovered on Day 1 under optimum conditions for the growth of Erwinia (90% RH, 27C), and also on Day 2 under fall/winter production conditions (60% RH, 27C). These specific time points had relatively large error bars indicating that human error in counting appropriate dilution plates or contaminated plates from improper handling techniques may have been possible factors in the elevated counts observed. Humidity and temperature are both important in the survival of Shigella on the surface of intact tomatoes. As seen with the Salmonella cocktail, the greatest recovery of cells was seen under conditions of higher humidity and temperature, or those of optimum conditions for the growth of Erwinia (90% RH, 27C).

PAGE 62

52 As with Salmonella and Shigella, Erwinia carotovora also decreased over time (Figure 4-7) on the intact tomato surface. The optimum conditions for growth of Erwinia (90% RH, 27C) allowed the most recovery from the intact tomato surface. The standard ripening room conditions (90% RH, 20C) and the fall/winter production conditions (60% RH, 27C) had the least recovered Erwinia over time. There was an unexpected increase seen on Day 2 under optimum conditions for the growth of Erwinia (90% RH, 27C). Additionally, the standard deviation on Day 2 under standard ripening room conditions was very large. One explanation for this event could be that one of the samples taken that day had a break in the skin that had not been seen before spot inoculation and had developed soft rot. This resulted in the growth of Erwinia on what originally appeared to be an intact tomato. Humidity and temperature plays an important role in the survival of Erwinia on intact tomato surfaces. At its optimum conditions, Erwinia was able to survive the best, whereas the lower humidity or lower temperature conditions significantly affected the survival of Erwinia after initial inoculation. The bacteria Salmonella, Shigella and Erwinia seem to survive but not proliferate on intact tomatoes under high humidity and room temperature (27C). The bacteria used in this research were observed to decline on the outer surface of tomatoes over time and the rate of reduction seemed to be related to both temperature and humidity. Growth of the three organisms on intact tomatoes was not seen. Foodborne bacteria are not capable of producing the enzymes necessary to breakdown the protective outer barriers on produce, which prevents nutrients from becoming available (Wells and Butterfield 1999). Although Erwinia does have the capability to produce these enzymes, unless there is a break in the tomato skin or other opening, it cannot produce soft rot (Toth et al. 2003).

PAGE 63

53 Salmonella was recovered from tomato surfaces in all three simulated environments and this indicates that the pathogen can survive on tomato surfaces and should be a concern in the fresh produce industry. Cross-contamination is possible if contaminated fruit are mixed in with other clean fruit, which may pose a threat to food safety. Recovery of Bacteria from Wounded Tomato Surfaces Shave wounded tomatoes were stored for three days in the same environmental conditions as mentioned previously. In accordance with findings by Wells and Butterfield (1997), the Salmonella cocktail was able to not only survive on the wounded tomato surface, but was capable of growing. The greatest amount of growth was seen under standard ripening room conditions (90% RH, 20C), closely followed by the optimum conditions for the growth of Erwinia (90% RH, 27C) (Figure 4-8). The slowest rate of growth occurred under fall/winter production season conditions (60% RH, 27C) showing that even at lower humidity, Salmonella is capable of not only surviving on the surface of wounded tomatoes, but growing. In agreement with Beattie and Lindow (1999), these findings suggest that a high relative humidity is imperative in the exponential growth of Salmonella on wounded tomato surfaces. Similar to Salmonella on shave-wounded tomatoes, the Shigella cocktail was able to survive and grow on the wounded tomato surface under higher humidity conditions (Figure 4-9). On Day 2 under optimum conditions for the growth of Erwinia (90% RH, 27C), there was a slight decrease in recovered cells, though the standard deviations seen were substantially large indicating other factors influencing growth and recovery. Unlike the Salmonella cocktail, Shigella was not recoverable at any amount under the fall/winter production season conditions (60% RH, 27C). A possible explanation for this is that the inoculum may desiccate to a point which would not support growth and/or recovery. In

PAGE 64

54 addition, at each subsequent day of the study, it was observed that the wound sites on the tomatoes kept at the lower humidity exhibited excessive desiccation and shriveling at the site. When inoculated by itself, Shigella requires a relatively high humidity to survive on wounded tomato surfaces. As expected, Erwinia carotovora was capable of survival and growth on wounded tomato surfaces at all environmental conditions (Figure 4-10). The greatest amount of growth occurred under their optimum condition (90% RH, 27C) which was not significantly greater than standard ripening room conditions (90% RH, 20C). The slowest rate of growth occurred under conditions of lower humidity, the fall/winter production conditions (60% RH, 27C). As seen with the wounded tomatoes for both the Salmonella and Shigella cocktails, at the lower humidity the wound sites exhibited some desiccation preventing the exponential growth seen at the higher humidity conditions. Although 20C is colder than what is preferential for the growth of Erwinia, under high humidity, the conditions remained favorable enough that the plant pathogen could survive and grow. The bacteria Salmonella and Erwinia are capable of growing and being recovered from wounded tomato surfaces at all three environmental conditions used. Shigella was able to proliferate at the higher humidity conditions, but not at a humidity of 60%. If there were a break in the surface of the tomato, there is a distinct possibility that a foodborne pathogen could survive and cause illness under the right conditions. Tomatoes with Salmonella and Shigella present would not show signs of being contaminated as opposed to soft rot seen with Erwinia. Since Erwinia produces a characteristic soft rot, it

PAGE 65

55 is possible for packinghouses to carefully observe the tomato fruit for soft rot and carry out culling practices to prevent the spread of this plant pathogen among the tomato bins. Recovery of Combined Bacteria from Wounded Tomato Surfaces In accordance with the findings of Wells and Butterfield (1997), Salmonella grew to higher levels in combination with Erwinia under fall/winter conditions and optimum conditions for Erwinia (Table 4-2; Table 4-3) when compared to Salmonella grown individually. The greatest amount of Salmonella growth was seen under optimum conditions for the growth of Erwinia (90% RH, 27C), while the greatest amount of growth for Erwinia was seen under standard ripening room conditions (90% RH, 20C) (Table 4-4). It can be determined from this research that the bacterial combination studies involving Salmonella and Erwinia did not grow at optimum levels under one specific environmental condition. Since both environmental conditions are acceptable for the growth of the bacteria, it may be possible that they are competing for nutrients and the available space to multiply. The surface pH of a tomato is approximately 4.5 (Guo et al. 2001). When Erwinia begins its soft rot pathogenesis on tomatoes the pH may increase slightly. An increase in pH on the tomato surface would be more favorable to Salmonella, which prefers a neutral pH, and could explain the greater amount of growth seen of Salmonella on wounded tomato surfaces when combined with Erwinia. A possible explanation for the lack of coordinated growth of both the Salmonella and Erwinia in combination on wounded tomato surfaces under each environmental condition could be that quorum sensing is taking place (Smith et al. 2003). Since the bacteria are competing for nutrients in a limited amount of space, it is possible that either Salmonella or Erwinia has reached a quorum and is entering survival mode until conditions become favorable again. If these studies were to continue for several more

PAGE 66

56 days, it would be possible to observe if any competition among bacteria existed, resulting in the death of one bacteria and growth of another. The Shigella and Erwinia combination studies all displayed significant growth over three days under all environmental conditions (Tables 4-5; 4-6; 4-7). The greatest amount of growth for both Shigella and Erwinia was seen at the 90% RH, 27C growth parameters. This was to be expected since Erwinia should thrive under those conditions, and Shigella is capable of growth under high humidity and temperature conditions as seen in previous studies. When inoculated alone, Shigella is not capable of survival on wounded tomato surfaces under fall/winter production season conditions (60% RH, 27C). By contrast, when combined with the soft rot plant pathogen Erwinia, Shigella was capable of survival as well as growth. Shigella was recovered on Day 0 under low humidity conditions when combined with Erwinia although there were no countable levels of Shigella on Day 1 and 2. Day 3 showed growth of Shigella at levels higher than the wounded tomatoes inoculated with Shigella alone at higher humidity conditions. It is possible that the Shigella was still present on the surface of the shave-wounded tomato but was at such low levels that it was below detection for standard plating techniques. Future studies could repeat these conditions utilizing a most probable number method (MPN) assay to determine if there are viable cells of Shigella present during Day 1 and 2. If the initial pH of the wounded tomato surface, approximately 4.5, was preventing the growth of Shigella, the enzymes produced during the soft rot pathogenesis of Erwinia can increase the pH to a more favorable level of 5.5. Although this is not a neutral pH which Shigella favors, there is less stress placed on the Shigella cells since the pH is slightly higher.

PAGE 67

57 Salmonella and Shigella may be able to grow to greater levels when Erwinia carotovora is present because the soft rot increases the surface area that these pathogens can colonize. Available water and damaged tomato cells resulting from tissue maceration can greatly increase the area that becomes permeated with fluid allowing for movement and multiplication of bacteria. Under normal circumstances, the relatively rigid tomato cell walls and definite intercellular spaces would limit bacterial motility specifically to areas with cell fluid present from the initial shave wound injury. It may be possible that bacteria can move on a moist surface if the intercellular spaces were congested with fluid resulting from soft rot. The opportunity for foodborne illness caused by Salmonella and Shigella on fresh produce, specifically tomatoes, can be greatly enhanced in the presence of the soft rot organism Erwinia carotovora. All three environmental conditions allowed for the survival and growth of the bacteria on the wounded tomato surface when in combination. To ensure food safety, it is important to remove any damaged fruit to prevent the spread of soft rot and the possibility of foodborne illness. It would be necessary to consider that if a tomato appears to have soft rot from Erwinia carotovora, it could possibly be contaminated with a foodborne pathogen such as Salmonella or Shigella, especially when the produce is held at higher temperature and humidity. Since Salmonella and Shigella have a relatively low infectious dose, it is important to inspect fresh produce for signs of soft rot and prevent the possibility of foodborne illness.

PAGE 68

CHAPTER 6 CONCLUSION All objectives of this study were completed. Growth rates of five rifampicin resistant Salmonella enterica serovars, two Shigella species and Erwinia carotovora were established and it was determined that appropriate cocktails could be made. Salmonella, Shigella and Erwinia were successfully recovered from intact tomatoes as well as shave-wounded tomatoes held at all three environmental conditions. It was observed that the individual microorganisms survived best on intact tomatoes held at optimum conditions for the growth of Erwinia (90% RH, 27C). The individual pathogens Salmonella and Shigella exhibited the most growth on shave wounded tomatoes under standard ripening room conditions (90% RH, 20C), while the Erwinia had the greatest amount of growth under its optimum conditions (90% RH, 27C). The lower humidity seen during the fall/winter season condition (60% RH, 27C) caused desiccation of the wounds on the shave-wounded fruit, resulting in lower recovery of the bacteria. In combined microorganism studies, Salmonella exhibited the greatest amount of growth with an increase of 4.81 log10 CFU/mL under optimum conditions for the growth of Erwinia (90% RH, 27C), while the greatest growth of Erwinia occurred under standard ripening room conditions (90% RH, 20C) with an increase of 6.26 log10 CFU/mL. Under optimum conditions for the growth of Erwinia, the log10 CFU/mL counts for Salmonella and Erwinia were relatively similar, 4.81 log10 CFU/mL and 4.94 log10 CFU/mL respectively, possibly showing an antagonistic relationship between the 58

PAGE 69

59 two bacteria at those environmental conditions. The least amount of growth occurred under fall/winter season conditions (60% RH, 27C) for both the Salmonella and Erwinia. These results show that relative humidity plays an important role in the growth of Salmonella and Erwinia. In combined bacteria studies involving Shigella and Erwinia, the optimum conditions for growth of Erwinia (90% RH, 27C) allowed for the most growth for both bacteria resulting in a 4.84 log10 CFU/mL increase of Shigella and a 6.55 log10 CFU/mL increase of Erwinia. The growth levels for Shigella and Erwinia were slightly less under standard ripening room conditions (90% RH, 20C). The least growth was observed under the fall/winter tomato season conditions (60% RH, 27C). These results show that relative humidity, and not necessarily temperature, plays an important role in the survival and growth of both Erwinia and Shigella on wounded tomato surfaces.

PAGE 70

APPENDIX A SALMONELLA STATISTICS Salmonella 60% RH, 27C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3.26669181 3.26669181 4.54 0.0658 Error 8 5.76128540 0.72016068 Corrected Total 9 9.02797721 R-Square Coeff Var Root MSE logct Mean 0.361841 66.53039 0.848623 1.275542 Source DF Type III SS Mean Square F Value Pr > F x 1 3.26669181 3.26669181 4.54 0.0658 t Tests (LSD) for logct NOTE: This test controls the Type I comparisonwise error rate, not the experimentwise error rate. Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0.720161 Critical Value of t 2.30600 Least Significant Difference 1.2377 Means with the same letter are not significantly different. t Grouping Mean N x A 1.8471 5 combo A A 0.7040 5 wounded 60

PAGE 71

61 ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.15543332 1.15543332 0.86 0.3800 Error 8 10.70946451 1.33868306 Corrected Total 9 11.86489783 R-Square Coeff Var Root MSE logct Mean 0.097382 92.19238 1.157015 1.255000 Source DF Type III SS Mean Square F Value Pr > F x 1 1.15543332 1.15543332 0.86 0.3800 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 1.338683 Critical Value of t 2.30600 Least Significant Difference 1.6874 t Grouping Mean N x A 1.5949 5 combo A A 0.9151 5 wounded ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10

PAGE 72

62 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 19.56421411 19.56421411 14.06 0.0056 Error 8 11.13356508 1.39169563 Corrected Total 9 30.69777918 R-Square Coeff Var Root MSE logct Mean 0.637317 41.83385 1.179702 2.819969 Source DF Type III SS Mean Square F Value Pr > F x 1 19.56421411 19.56421411 14.06 0.0056 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 1.391696 Critical Value of t 2.30600 Least Significant Difference 1.7205 t Grouping Mean N x A 4.2187 5 combo B 1.4212 5 wounded ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 16.41247704 16.41247704 9.94 0.0136 Error 8 13.21387226 1.65173403 Corrected Total 9 29.62634930 R-Square Coeff Var Root MSE logct Mean 0.553982 39.22735 1.285198 3.276281 Source DF Type III SS Mean Square F Value Pr > F x 1 16.41247704 16.41247704 9.94 0.0136

PAGE 73

63 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 1.651734 Critical Value of t 2.30600 Least Significant Difference 1.8744 t Grouping Mean N x A 4.5574 5 combo B 1.9952 5 wounded Salmonella 90% RH, 20C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 2 combo wounded Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.88614749 1.88614749 4.91 0.0911 Error 4 1.53811471 0.38452868 Corrected Total 5 3.42426220 R-Square Coeff Var Root MSE logct Mean 0.550819 41.98408 0.620104 1.476997 Source DF Type III SS Mean Square F Value Pr > F x 1 1.88614749 1.88614749 4.91 0.0911 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.384529 Critical Value of t 2.77645 Least Significant Difference 1.4057 t Grouping Mean N x A 2.0377 3 wounded A A 0.9163 3 combo

PAGE 74

64 ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 2 combo wounded Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.98605237 1.98605237 3.74 0.1254 Error 4 2.12624746 0.53156186 Corrected Total 5 4.11229983 R-Square Coeff Var Root MSE logct Mean 0.482954 17.14065 0.729083 4.253530 Source DF Type III SS Mean Square F Value Pr > F x 1 1.98605237 1.98605237 3.74 0.1254 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.531562 Critical Value of t 2.77645 Least Significant Difference 1.6528 t Grouping Mean N x A 4.8289 3 combo A A 3.6782 3 wounded ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 2 combo wounded Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct

PAGE 75

65 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.43025008 1.43025008 4.67 0.0969 Error 4 1.22623808 0.30655952 Corrected Total 5 2.65648816 R-Square Coeff Var Root MSE logct Mean 0.538399 10.98203 0.553678 5.041675 Source DF Type III SS Mean Square F Value Pr > F x 1 1.43025008 1.43025008 4.67 0.0969 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.30656 Critical Value of t 2.77645 Least Significant Difference 1.2552 t Grouping Mean N x A 5.5299 3 wounded A A 4.5534 3 combo ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 2 combo wounded Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.37074825 1.37074825 8.04 0.0471 Error 4 0.68177465 0.17044366 Corrected Total 5 2.05252290 R-Square Coeff Var Root MSE logct Mean 0.667836 7.432287 0.412848 5.554794 Source DF Type III SS Mean Square F Value Pr > F x 1 1.37074825 1.37074825 8.04 0.0471

PAGE 76

66 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.170444 Critical Value of t 2.77645 Least Significant Difference 0.9359 t Grouping Mean N x A 6.0328 3 wounded B 5.0768 3 combo Salmonella 90% RH 27C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 8 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.51312168 0.51312168 6.72 0.0411 Error 6 0.45827522 0.07637920 Corrected Total 7 0.97139691 R-Square Coeff Var Root MSE logct Mean 0.528231 15.51868 0.276368 1.780872 Source DF Type III SS Mean Square F Value Pr > F x 1 0.51312168 0.51312168 6.72 0.0411 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 6 Error Mean Square 0.076379 Critical Value of t 2.44691 Least Significant Difference 0.4939 Harmonic Mean of Cell Sizes 3.75 t Grouping Mean N x A 1.9770 5 wounded B 1.4539 3 combo

PAGE 77

67 ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 2.05174567 2.05174567 3.48 0.0993 Error 8 4.72285687 0.59035711 Corrected Total 9 6.77460254 R-Square Coeff Var Root MSE logct Mean 0.302858 16.25309 0.768347 4.727392 Source DF Type III SS Mean Square F Value Pr > F x 1 2.05174567 2.05174567 3.48 0.0993 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0.590357 Critical Value of t 2.30600 Least Significant Difference 1.1206 t Grouping Mean N x A 5.1804 5 combo A A 4.2744 5 wounded ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10

PAGE 78

68 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.30312206 0.30312206 0.13 0.7278 Error 8 18.66443911 2.33305489 Corrected Total 9 18.96756116 R-Square Coeff Var Root MSE logct Mean 0.015981 28.44271 1.527434 5.370213 Source DF Type III SS Mean Square F Value Pr > F x 1 0.30312206 0.30312206 0.13 0.7278 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 2.333055 Critical Value of t 2.30600 Least Significant Difference 2.2277 t Grouping Mean N x A 5.5443 5 wounded A A 5.1961 5 combo ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 2.08961791 2.08961791 2.73 0.1370 Error 8 6.12057870 0.76507234 Corrected Total 9 8.21019661 R-Square Coeff Var Root MSE logct Mean 0.254515 15.13173 0.874684 5.780462

PAGE 79

69 Source DF Type III SS Mean Square F Value Pr > F x 1 2.08961791 2.08961791 2.73 0.1370 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0.765072 Critical Value of t 2.30600 Least Significant Difference 1.2757 t Grouping Mean N x A 6.2376 5 combo A A 5.3233 5 wounded

PAGE 80

APPENDIX B SHIGELLA STATISTICS Shigella 60% RH, 27C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3.52569825 3.52569825 161.62 <.0001 Error 8 0.17452325 0.02181541 Corrected Total 9 3.70022149 R-Square Coeff Var Root MSE logct Mean 0.952834 24.87477 0.147700 0.593776 Source DF Type III SS Mean Square F Value Pr > F x 1 3.52569825 3.52569825 161.62 <.0001 t Tests (LSD) for logct NOTE: This test controls the Type I comparisonwise error rate, not the experimentwise error rate. Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0.021815 Critical Value of t 2.30600 Least Significant Difference 0.2154 Means with the same letter are not significantly different. 70

PAGE 81

71 t Grouping Mean N x A 1.18755 5 combo B 0.00000 5 wounded ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0 0 . Error 8 0 0 Corrected Total 9 0 R-Square Coeff Var Root MSE logct Mean 0.000000 0 0 Source DF Type III SS Mean Square F Value Pr > F x 1 0 0 . t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0 Critical Value of t 2.30600 Least Significant Difference 0 t Grouping Mean N x A 0 5 combo A A 0 5 wounded ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 2 combo wounded

PAGE 82

72 Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0 0 . Error 8 0 0 Corrected Total 9 0 R-Square Coeff Var Root MSE logct Mean 0.000000 0 0 Source DF Type III SS Mean Square F Value Pr > F x 1 0 0 . t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0 Critical Value of t 2.30600 Least Significant Difference 0 t Grouping Mean N x A 0 5 combo A A 0 5 wounded ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 64.56156527 64.56156527 230.78 <.0001 Error 8 2.23807082 0.27975885 Corrected Total 9 66.79963610

PAGE 83

73 R-Square Coeff Var Root MSE logct Mean 0.966496 20.81636 0.528922 2.540897 Source DF Type III SS Mean Square F Value Pr > F x 1 64.56156527 64.56156527 230.78 <.0001 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0.279759 Critical Value of t 2.30600 Least Significant Difference 0.7714 t Grouping Mean N x A 5.0818 5 combo B 0.0000 5 wounded Shigella 90% RH, 20C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 2 combo wounded Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 2.49520030 2.49520030 10.75 0.0305 Error 4 0.92843045 0.23210761 Corrected Total 5 3.42363076 R-Square Coeff Var Root MSE logct Mean 0.728817 48.56568 0.481775 0.992008 Source DF Type III SS Mean Square F Value Pr > F x 1 2.49520030 2.49520030 10.75 0.0305 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.232108 Critical Value of t 2.77645 Least Significant Difference 1.0922

PAGE 84

74 t Grouping Mean N x A 1.6369 3 wounded B 0.3471 3 combo ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 2 combo wounded Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.00101221 0.00101221 0.06 0.8241 Error 4 0.07191014 0.01797753 Corrected Total 5 0.07292235 R-Square Coeff Var Root MSE logct Mean 0.013881 5.099104 0.134080 2.629488 Source DF Type III SS Mean Square F Value Pr > F x 1 0.00101221 0.00101221 0.06 0.8241 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.017978 Critical Value of t 2.77645 Least Significant Difference 0.304 t Grouping Mean N x A 2.6425 3 combo A A 2.6165 3 wounded ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 2 combo wounded

PAGE 85

75 Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.03769405 0.03769405 1.59 0.2761 Error 4 0.09492378 0.02373094 Corrected Total 5 0.13261782 R-Square Coeff Var Root MSE logct Mean 0.284231 5.098507 0.154049 3.021444 Source DF Type III SS Mean Square F Value Pr > F x 1 0.03769405 0.03769405 1.59 0.2761 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.023731 Critical Value of t 2.77645 Least Significant Difference 0.3492 t Grouping Mean N x A 3.1007 3 wounded A A 2.9422 3 combo ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 2 combo wounded Number of Observations Read 6 Number of Observations Used 6 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.13916497 0.13916497 0.52 0.5096 Error 4 1.06430764 0.26607691 Corrected Total 5 1.20347261

PAGE 86

76 R-Square Coeff Var Root MSE logct Mean 0.115636 13.45912 0.515826 3.832543 Source DF Type III SS Mean Square F Value Pr > F x 1 0.13916497 0.13916497 0.52 0.5096 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 4 Error Mean Square 0.266077 Critical Value of t 2.77645 Least Significant Difference 1.1694 t Grouping Mean N x A 3.9848 3 wounded A A 3.6802 3 combo Shigella 90% RH, 27C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 8 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.87514338 1.87514338 10.69 0.0170 Error 6 1.05244211 0.17540702 Corrected Total 7 2.92758549 R-Square Coeff Var Root MSE logct Mean 0.640508 31.74568 0.418816 1.319286 Source DF Type III SS Mean Square F Value Pr > F x 1 1.87514338 1.87514338 10.69 0.0170

PAGE 87

77 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 6 Error Mean Square 0.175407 Critical Value of t 2.44691 Least Significant Difference 0.7484 Harmonic Mean of Cell Sizes 3.75 t Grouping Mean N x A 1.6943 5 wounded B 0.6943 3 combo ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 16.55253730 16.55253730 15.58 0.0043 Error 8 8.50111961 1.06263995 Corrected Total 9 25.05365691 R-Square Coeff Var Root MSE logct Mean 0.660683 46.69006 1.030844 2.207845 Source DF Type III SS Mean Square F Value Pr > F x 1 16.55253730 16.55253730 15.58 0.0043 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 1.06264 Critical Value of t 2.30600 Least Significant Difference 1.5034 t Grouping Mean N x A 3.4944 5 wounded B 0.9213 5 combo

PAGE 88

78 ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.12107542 0.12107542 0.03 0.8566 Error 8 27.82375321 3.47796915 Corrected Total 9 27.94482864 R-Square Coeff Var Root MSE logct Mean 0.004333 60.00267 1.864931 3.108081 Source DF Type III SS Mean Square F Value Pr > F x 1 0.12107542 0.12107542 0.03 0.8566 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 3.477969 Critical Value of t 2.30600 Least Significant Difference 2.7199 t Grouping Mean N x A 3.218 5 combo A A 2.998 5 wounded ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 2 combo wounded Number of Observations Read 10 Number of Observations Used 10 Dependent Variable: logct

PAGE 89

79 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 16.05186754 16.05186754 13.91 0.0058 Error 8 9.23255826 1.15406978 Corrected Total 9 25.28442580 R-Square Coeff Var Root MSE logct Mean 0.634852 25.32880 1.074276 4.241323 Source DF Type III SS Mean Square F Value Pr > F x 1 16.05186754 16.05186754 13.91 0.0058 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 1.15407 Critical Value of t 2.30600 Least Significant Difference 1.5668 t Grouping Mean N x A 5.5083 5 combo B 2.9744 5 wounded

PAGE 90

APPENDIX C ERWINIA STATISTICS Erwinia 60% RH, 27C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 15 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 10.50799005 5.25399502 19.08 0.0002 Error 12 3.30449643 0.27537470 Corrected Total 14 13.81248648 R-Square Coeff Var Root MSE logct Mean 0.760760 36.37229 0.524762 1.442751 Source DF Type III SS Mean Square F Value Pr > F x 2 10.50799005 5.25399502 19.08 0.0002 t Tests (LSD) for logct NOTE: This test controls the Type I comparisonwise error rate, not the experimentwise error rate. Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 0.275375 Critical Value of t 2.17881 Least Significant Difference 0.7231 Means with the same letter are not significantly different. 80

PAGE 91

81 t Grouping Mean N x A 2.1293 5 combosal A A 1.9345 5 comboshi B 0.2644 5 wounded ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 15 Sum of Source DF Squares Mean Square F Value Pr > F Model 2 1.65440723 0.82720361 0.80 0.4726 Error 12 12.43466317 1.03622193 Corrected Total 14 14.08907040 R-Square Coeff Var Root MSE logct Mean 0.117425 46.16068 1.017950 2.205231 Source DF Type III SS Mean Square F Value Pr > F x 2 1.65440723 0.82720361 0.80 0.4726 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 1.036222 Critical Value of t 2.17881 Least Significant Difference 1.4027 t Grouping Mean N x A 2.6748 5 combosal A A 1.9767 5 wounded A A 1.9641 5 comboshi

PAGE 92

82 ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 15 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 48.63180315 24.31590157 8.61 0.0048 Error 12 33.90544483 2.82545374 Corrected Total 14 82.53724798 R-Square Coeff Var Root MSE logct Mean 0.589210 48.30976 1.680909 3.479439 Source DF Type I SS Mean Square F Value Pr > F x 2 48.63180315 24.31590157 8.61 0.0048 Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 2.825454 Critical Value of t 2.17881 Least Significant Difference 2.3163 t Grouping Mean N x A 6.000 5 combosal B 2.531 5 comboshi B B 1.907 5 wounded ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 15

PAGE 93

83 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 36.41360097 18.20680049 7.45 0.0079 Error 12 29.32981256 2.44415105 Corrected Total 14 65.74341353 R-Square Coeff Var Root MSE logct Mean 0.553875 30.94536 1.563378 5.052060 Source DF Type III SS Mean Square F Value Pr > F x 2 36.41360097 18.20680049 7.45 0.0079 Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 2.444151 Critical Value of t 2.17881 Least Significant Difference 2.1543 t Grouping Mean N x A 6.2014 5 comboshi A A 6.1055 5 combosal B 2.8493 5 wounded Erwinia 90% RH, 20C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 3 combosal comboshi wounded Number of Observations Read 9 Number of Observations Used 9 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 0.14695946 0.07347973 0.34 0.7264 Error 6 1.30692086 0.21782014 Corrected Total 8 1.45388032 R-Square Coeff Var Root MSE logct Mean 0.101081 24.06714 0.466712 1.939208

PAGE 94

84 Source DF Type I SS Mean Square F Value Pr > F x 2 0.14695946 0.07347973 0.34 0.7264 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 6 Error Mean Square 0.21782 Critical Value of t 2.44691 Least Significant Difference 0.9324 t Grouping Mean N x A 2.1150 3 wounded A A 1.8876 3 combosal A A 1.8151 3 comboshi ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 3 combosal comboshi wounded Number of Observations Read 9 Number of Observations Used 9 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 16.51690575 8.25845287 21.09 0.0019 Error 6 2.34960666 0.39160111 Corrected Total 8 18.86651240 R-Square Coeff Var Root MSE logct Mean 0.875462 13.04686 0.625780 4.796407 Source DF Type III SS Mean Square F Value Pr > F x 2 16.51690575 8.25845287 21.09 0.0019 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 6 Error Mean Square 0.391601 Critical Value of t 2.44691 Least Significant Difference 1.2502

PAGE 95

85 t Grouping Mean N x A 6.0351 3 wounded A A 5.4428 3 combosal B 2.9113 3 comboshi ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 3 combosal comboshi wounded Number of Observations Read 9 Number of Observations Used 9 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 0.80993932 0.40496966 2.43 0.1686 Error 6 0.99978794 0.16663132 Corrected Total 8 1.80972725 R-Square Coeff Var Root MSE logct Mean 0.447548 6.146878 0.408205 6.640851 Source DF Type I SS Mean Square F Value Pr > F x 2 0.80993932 0.40496966 2.43 0.1686 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 6 Error Mean Square 0.166631 Critical Value of t 2.44691 Least Significant Difference 0.8156 t Grouping Mean N x A 7.0541 3 wounded A A 6.5173 3 comboshi A A 6.3512 3 combosal

PAGE 96

86 ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 3 combosal comboshi wounded Number of Observations Read 9 Number of Observations Used 9 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 0.75602073 0.37801037 25.93 0.0011 Error 6 0.08746506 0.01457751 Corrected Total 8 0.84348579 R-Square Coeff Var Root MSE logct Mean 0.896305 1.580899 0.120737 7.637260 Source DF Type I SS Mean Square F Value Pr > F x 2 0.75602073 0.37801037 25.93 0.0011 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 6 Error Mean Square 0.014578 Critical Value of t 2.44691 Least Significant Difference 0.2412 t Grouping Mean N x A 8.03550 3 wounded B 7.52213 3 comboshi B B 7.35415 3 combosal Erwinia 90% RH, 27C ----------------------------------day=0 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 0 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 11

PAGE 97

87 Sum of Source DF Squares Mean Square F Value Pr > F Model 2 2.75162122 1.37581061 5.39 0.0329 Error 8 2.04224723 0.25528090 Corrected Total 10 4.79386845 R-Square Coeff Var Root MSE logct Mean 0.573988 29.09062 0.505253 1.736825 Source DF Type I SS Mean Square F Value Pr > F x 2 2.75162122 1.37581061 5.39 0.0329 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 8 Error Mean Square 0.255281 Critical Value of t 2.30600 Comparisons significant at the 0.05 level are indicated by ***. Difference x Between 95% Confidence Comparison Means Limits wounded combosal 0.5524 -0.2985 1.4032 wounded comboshi 1.2066 0.3557 2.0575 *** combosal wounded -0.5524 -1.4032 0.2985 combosal comboshi 0.6543 -0.2971 1.6056 comboshi wounded -1.2066 -2.0575 -0.3557 *** comboshi combosal -0.6543 -1.6056 0.2971 ----------------------------------day=1 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 1 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 15 Dependent Variable: logct Sum of Source DF Squares Mean Square F Value Pr > F Model 2 10.32400292 5.16200146 10.85 0.0020 Error 12 5.71129079 0.47594090 Corrected Total 14 16.03529371 R-Square Coeff Var Root MSE logct Mean 0.643830 13.31123 0.689885 5.182726

PAGE 98

88 Source DF Type III SS Mean Square F Value Pr > F x 2 10.32400292 5.16200146 10.85 0.0020 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 0.475941 Critical Value of t 2.17881 Least Significant Difference 0.9507 t Grouping Mean N x A 5.9190 5 wounded A A 5.6056 5 combosal B 4.0235 5 comboshi ----------------------------------day=2 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 2 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 15 Sum of Source DF Squares Mean Square F Value Pr > F Model 2 9.31016860 4.65508430 5.88 0.0166 Error 12 9.49263861 0.79105322 Corrected Total 14 18.80280722 R-Square Coeff Var Root MSE logct Mean 0.495148 13.03674 0.889412 6.822349 Source DF Type III SS Mean Square F Value Pr > F x 2 9.31016860 4.65508430 5.88 0.0166 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 0.791053 Critical Value of t 2.17881 Least Significant Difference 1.2256

PAGE 99

89 t Grouping Mean N x A 7.9329 5 wounded B 6.3450 5 combosal B B 6.1892 5 comboshi ----------------------------------day=3 -----------------------------------The GLM Procedure Class Level Information Class Levels Values day 1 3 x 3 combosal comboshi wounded Number of Observations Read 15 Number of Observations Used 15 Sum of Source DF Squares Mean Square F Value Pr > F Model 2 12.33484301 6.16742151 34.59 <.0001 Error 12 2.13991544 0.17832629 Corrected Total 14 14.47475845 R-Square Coeff Var Root MSE logct Mean 0.852162 5.521948 0.422287 7.647427 Source DF Type III SS Mean Square F Value Pr > F x 2 12.33484301 6.16742151 34.59 <.0001 t Tests (LSD) for logct Alpha 0.05 Error Degrees of Freedom 12 Error Mean Square 0.178326 Critical Value of t 2.17881 Least Significant Difference 0.5819 t Grouping Mean N x A 8.8046 5 wounded B 7.5475 5 comboshi C 6.5901 5 combosal

PAGE 100

LIST OF REFERENCES Allen, R.L. 2003. A Recovery Study of Salmonella spp. from the Surfaces of Tomatoes and Packing Line Materials; Masters Thesis [electronic resource]. Gainesville, FL. University of Florida. Altekurse, S.F., M.L. Cohen and D.L. Swerdlow. 1997. Emerging Foodborne Diseases. Emerg. Inf. Dis. 3(3):285-293. Asplund, K. and E. Nurmi. 1991. The Growth of Salmonella in Tomatoes. Int. J. Food Micro. 13:177-182. Aysan, Y., A. Karatas and O. Cinar. 2003. Biological Control of Bacterial Stem Rot Caused by Erwinia chrysanthemi on Tomato. Crop Prot. 22:807-811. Bagamboula, C.F., M. Uyttendaele and J. Debevere. 2002. Acid Tolerance of Shigella sonnei and Shigella flexneri. J. App. Microbiol. 93(3):479-486. Baker, R.C., R.A. Qureshi and J.H. Hotchkiss. 1986. Effect of Elevated Level of Carbon Dioxide on the Growth of Spoilage and Pathogenic Bacteria. Poult. Sci. 65:729-737. Barras, F., F. van Gijsegem and A.K. Chatterjee. 1994. Extracellular Enzymes and Pathogenesis of the Soft Rot Erwinia. Ann. Rev. Phytopathol. 32:201-234. Bartz, J.A. and R.K. Showalter. 1981. Infiltration of Tomatoes by Aqueous Bacterial Suspension. Phytopath. 71:515-518. Beattie, G.A. and S.E. Lindow. 1999. Bacterial colonization of leaves: A spectrum of strategies. Phytopathol. 89(5):353-9. Beuchat, L.R. 1995. Pathogenic Microorganisms Associated with Fresh Produce. J. Food Prot. 59(2):204-216. Beuchat, L.R. 2002. Ecological Factors Influencing Survival and Growth of Human Pathogens on Raw Fruits and Vegetables. Microbes Infect. 4:413-423. Beuchat, L.R., A.J. Scouten, R.I. Allen and R.S. Hussey. 2003. Potential of a Plant-Parasitic Nematode to Facilitate Internal Contamination of Tomato Plants by Salmonella. J. Food Prot. 66(8):1459-1461. 90

PAGE 101

91 Blostein, J. 1991. An Outbreak of Salmonella javiana Associated with Consumption of Watermelon. J. Environ. Health. 56:29-31. Brackett, R.E. 1992. Shelf Stability and Safety of Fresh Produce as Influenced by Sanitation and Disinfection. J. Food Prot. 55(10):808-814. Brackett, R.E. 1999. Incidence, Contributing Factors, and Control of Bacterial Pathogens in Produce. Postharvest Bio. and Tech. 15(1999):305-311. Buchanan, R.E. and N.E. Gibbons. 1974. Gram-Negative Facultatively Anaerobic Rods. In: Carpenter KP (Ed.) Bergeys Manual of Determinative Bacteriology. 8th Ed. p. 318. Baltimore: The Williams and Wilkins Company. Burnett, A.B. and L.R. Beuchat. 2001. Comparison of Sample Preparation Methods for Recovering Salmonella from Raw Fruits, Vegetables, and Herbs. J. Food Prot. 64(10):1459-1465. Centers for Disease Control and Prevention. 1979. Salmonella oranienburg Gastroenteritis Associated with Consumption of Precut Watermelons. Morbid. Mortal. Weekly Rep. 28:522-523. Centers for Disease Control and Prevention. 1991. Multi-State Outbreak of Salmonella poona Infections United States and Canada. Morb. Mortal. Weekly. Rep. 40:549-552. Centers for Disease Control and Prevention. 1993. Multi-State Outbreak of Salmonella serotype Montevideo Infections. EPI-AID, p. 93-97. Centers for Disease Control and Prevention, Atlanta, GA. Centers for Disease Control and Prevention. 2002. Outbreak of Salmonella serotype Javiana Infections Orlando, Florida, June 2002. Morb. Mortal. Weekly. Rep. 51:683-684. Centers for Disease Control and Prevention. Salmonellosis. 2004a. CDC. http://www.cdc.gov/ncidod/dbmd/diseaseinfo/salmonellosis_g.htm. Accessed 2005 June 22. Centers for Disease Control and Prevention. Vegetable of the Month: Tomatoes. 2004b. CDC. http://www.cdc.gov/nccdphp/dnpa/5aday/month/pdfs/Tomatoes.pdf Accessed 2005 June 22. Center for Food Safety and Applied Nutrition. Analysis and evaluation of preventative control measures for the control and reduction/elimination of microbial hazards on fresh and fresh-cut produce. 2001. FDA. http://www.cfsan.fda.gov/~comm Accessed 2005 June 22. Cox, J. 2000a. Introduction. In: Robinson, R.K., C.A. Batt, and P.D. Patel (Ed.) Encyclopedia of Food Microbiology. Vol. 3, p. 1928. California: Academic Press.

PAGE 102

92 Cummings, K., E. Barrett, J.C. Mohle-Boetani, J.T. Brooks, A. Fiore, K. Komatsu, S.B. Werner and L. Slutsker. 2001. A Multistate Outbreak of Salmonella enterica Serotype Baildon Associated with Domestic Raw Tomatoes. Emer. Infec. Dis. 7:1046-1048. Droffner, M.L. and N. Yamamoto. 1992. Procedure for Isolation of Eschericia, Salmonella, and Pseudomonas Mutants Capable of Growth at Refractory Temperature of 54C. J. Micro. Meth. 14:201-206. DuPont, H.L. Shigella species. 1990. In: Mandell, G.L. R.G. Douglas and J.E. Bennett, (eds). Principles and Practice of Infectious Diseases. 3rd ed. New York: Churchhill Livingstone; 1990:1716-1722. Food and Drug Administration. FDA Survey of Imported Fresh Produce: FY 1999 Field Assignment. 2001. FDA. http://www.cfsan.fda.gov/~dms/prodsur6.html Accessed 2005 June 22. Food and Drug Administration. FDA Survey of Domestic Fresh Produce: FY 2000/2001 Field Assignemtn. 2003. FDA. http://www.cfsan.fda.gov/~dms/prodsu10.html Accessed 2005 June 22. Food and Drug Administration. Salmonella spp. 2005. FDA. http://www.cfsan.fda.gov/~mow/chap1.html Accessed 2005 June 23. Fray, R.G. 2001. Altering Pant-Microbe Interaction through Artificially Manipulating Bacterial Quorum Sensing. Annals Bot. 89:245-253. Guo, X., J. Chen, R.E. Bracket and L.R. Beuchat. 2001. Survival of Salmonella on and in Tomato Plants from the Time of Inoculation and Flowering and Early States of Fruit Development through Fruit Ripening. App. Envir. Micro. 67(10):4760-4764. Harris, L.J., D. Zagory and J.R. Gorny. 2002. Safety Factors. Pp.301-313. In: Adel A. Kader (Ed.) 3rd ed., Postharvest Technology of Horticultural Crops. University of California, Davis, Agriculture and Natural Resources, Publication 3311. Hedberg, C.W., F.J. Angulo, K.E. White, C.W. Langkop, W.L. Schell, M.G. Stobierski, A. Schuchat, J.M. Besser, S. Dietrich, L. Helsel, P.M. Griffin, J.W. McFarland and M.T. Osterholm. 1999. Outbreaks of Salmonellosis Associated with Eating Uncooked Tomatoes: Implications for Public Health. Epidemol. Infect. 122:385-393. Hedberg, C.W., K.L. Macdonald and M.T. Osterholm. 1994. Changing Epidemiology of Foodborne Disease: a Minnesota Perspective. Clin. Infect. Dis. 18:671-682. Holley, R.A. and M. Proulx. 1986. Use of Egg Wash pH to Prevent Survival of Salmonella at Moderate Temperatures. Poult. Sci. 65:922-928.

PAGE 103

93 Ibarra-Sanchez, L.S., S. Alvarado-Casillas, M.O. Rodriguez-Garcia, N.E. Martinez-Gonzales and A. Castillo. 2004. Internalization of Bacterial Pathogens in Tomatoes and Their Control by Selected Chemicals. J. Food Prot. 67(7):1353-1358. ICMSF (2002). Microorganisms in Foods 7. Microbiological Testing in Food Safety Management. New York, NY: Kluwer Academic/ Plenum Publishers. Kaspar, C.W. and M.L. Tamplin. 1993. Effects of Temperature and Salinity on the Survival of Vibrio vulnificus in Seawater and Shellfish. App. Envir. Micro. 59(8): 2425-2429. Kelman, A. 1956. Factors Influencing Viability and Variation in Cultures of Pseudomonas solanaearum. Phytopath. 46:16-17. Lampel, K.A., R.C. Sandlin and S. Formal. Shigella. In: Robinson, R.K., C.A. Batt, and P.D. Patel (Ed.) Encyclopedia of Food Microbiology. Vol. 3. p. 2015. California:Academic Press. Liao, C.H. and G.M. Sapers. 1999. Influence of Soft Rot Bacteria on Growth of Listeria monocytogenes on Potato Tuber Slices. J. Food Prot. 62(4):343-348. Liao, C.H. and L.M. Shollenberger. 2003. Survivability and Long-Term Preservation of Bacteria in Water and in Phosphate-Buffered Saline. Letters in App. Micro. 37(1):45-50. Lucier, G., B.H. Lin, J. Allshouse and L.S. Kantor. 2000. Factors Affecting Tomato Consumption in the United States. Economic Research Service/ USDA. VGS-282. Manefield, M., M. Welch, M. Givskov, G.P.C. Salmond and S. Kjelleberg. 2001. Halogenated Furanones from the Red Alga, Delisea pulchra, Inhibit Carbapenem Antibiotic Synthesis and Exoenzyme Virulence Factor Production in the Phytopathogen Erwinia carotovora. FEMS Micro. Letters. 205(2001):131-138. Mead, P.S., L. Slutsker, V. Dietz, L.F. McCaig, J.S. Bresee, C. Sharpivo, P.M. Griffin and R.V Tauxe. 1999. Food-Related Illness and Death in the United States. Emerg. Infect. Dis. 5:607-625. Molina, L., F. Constantinescu, L. Michel, C. Reimmann, B. Duffy and G Defago. 2003. Degradation of Pathogen Quorum-Sensing Molecules by Soil Bacteria: a Preventive and Curative Biological Control Mechanism. FEMS Micro. Eco. 45(2003):71-81. Nguyen, H.A., J. Kaneko and Y. Kamio. 2002. Temperature Dependent Production of Carotovoricin Er and Pectin Lyase in Phytopathogenic Erwinia carotovora subsp carotovora Er. Biosci. Biotech. Biochem. 66(2):444-447. Novak, J.S., G.M. Sapers and V.K. Juneja. 2003. Microbial Safety of Minimally Processed Foods. Boca Raton, FL:CRC Press, 267.

PAGE 104

94 Olson, S.M., D.N. Maynard, G.J. Hochmuth, C.S. Vavrina, W.M. Stall, T.A. Kucharek, S.E. Webb, T.G. Taylor, S.A. Smith and E.H. Simonne. 2004. Vegetable Production Handbook: Tomato Production in Florida, HS739, University of Florida, Gainesville, 32611. p. 301-316. Prombelon, M.C.M. 2002. Potato Diseases Caused by Soft Rot Erwinias. Ann. Rev. Phytopathol. 18:361-387. Prombelon, M.C.M. and A. Kelman. 1980. Ecology of Soft Rot Erwinias. Ann. Rev. Phytopathol. 18:361-387. Prombelon, M.C.M., V.M. Lumb and L.J. Hyman. 1987a. A Rapid Method to Identify and Quantify Soft Rot Erwinias on Seed Potato Tubers. EPPO Bull. 17:25-35. Prombelon, M.C.M., V.M. Lumb and D. Zutra. 1987b. Pathogenicity of Soft Rot Erwinias to Potato Plants in Scotland and Israel. J. Appl. Bacteriol. 63:73-84. Prombelon, M.C.M. and G.P.C. Salmond. 1995. Bacterial Soft Rots: In: Pathogenesis and Host Specificity in Plant Diseases, Vol. 1. Prokaryotes (Singh, U.S., Singh, R.P. and K. Kohmoto eds) p. 1-20. Oxford, U.K.: Pergamon. Ries, A.A., S. Zaza, C. Langkop, R.V. Tauxe and P.A. Balke. 1990. A Multistate Outbreak of Salmonella chester Linked to Imported Cantaloupe. Abst. 915, p. 238. Prog. Abst. 30th Interscience Conf. Antimicrob. Agents Chemother., American Society of Microbiology, Washington, D.C. Samelis, J. and J.N. Sofos. 2003. Microbial Stress Adaptation and Food Safety: Strategies to Control Stress-Adapted Pathogens. p. 304. In: A.E. Yousef and V.K. Juneja (Ed.) 1st ed., CRC Press. Sargent S. 1998. Handling Florida Vegetables-Tomato. SS-VEC-928, University of Florida, Gainesville, 32611. Smith, J.L. 1987. Shigella as a Foodborne Pathogen. J. Food Prot. 50:788-801. Smith, J.L., P.M. Fratamico and J.S. Novak. 2003. Quorum Sensing: A Primer for Food Microbiologists. J. Food Prot. 67(5):1053-1070. Tauxe, R.V., H. Kruse, C. Hedburg, M. Potter, J. Madden and K. Wachsmuth. 1997. Microbial Hazards and Emerging Issues Associated with Produce, a Preliminary Report to the National Advisory Committee on Microbial Criteria for Foods. J. Food Prot. 60(11):1400-1408. Thomson, N.R., J.D. Thomas and G.P.C. Salmond. 1999. Virulence Determinants in the Bacterial Phytopathogen Erwinia. Meth. Microbiol. 29:347-426. Toth, I.K., K.S. Bell, M.C. Holeva and P.R.J. Birch. 2003. Soft Rot Erwiniae: from Genes to Genomes. Mol. Plant Path. 4(1):17-30.

PAGE 105

95 United States Department of Agriculture. United States Standards for Grades of Fresh Tomatoes. 1997. USDA. http://www.ams.usda.gov/standards/tomatfrh.pdf Accessed 2005 June 22. United States Department of Agriculture. Tomatoes, Red, Ripe, Raw, June through October Average. 2004. USDA. http://www.nal.usda.gov/fnic/foodcomp/search/index.html Accessed 2005 June 22. Wade, W.N. and L.R. Beuchat. 2003. Metabiosis of Proteolytic Molds and Salmonella in Raw, Ripe Tomatoes. J. App. Microbiol. 95(3):437-450. Wells, J.M. and J.E. Butterfield. 1997. Salmonella Contamination Associated with Bacterial Soft Rot of Fesh Fruits and Vegetables in the Marketplace. Plant Dis. 81:867-872. Wells, J.M. and J.E. Butterfield. 1999. Incidence of Salmonella on Fresh Fruits and Vegetables Affected by Fungal Rots or Physical Injury. Plant Dis. 83:722-726. Wilson, C.L. and S. Droby. 2001. Microbial Food Contamination. p. 188. Boca Raton, FL: CRC Press LLC. Yoon, Y., J.D. Stopforth, P.A. Kendall and J.N. Sofos. 2004. Inactivation of Salmonella During Drying and Storage of Roma Tomatoes Exposed to Predrying Treatments Including Peeling, Blanching and Dipping in Organic Acid Solutions. J. Food Prot. 67(7):1344-1352. Zhuang, R.Y., L.R. Beuchat and F.J. Angulo. 1995. Fate of Salmonella Montevideo on and in Raw Tomatoes as Affected by Temperature and Treatment with Chlorine. J. Food Prot. 61:2127-2131.

PAGE 106

BIOGRAPHICAL SKETCH Jennifer Ann Joy was born an only child to John and Dolores Joy on December 10, 1982 in Deer Park, New York. She attended Deer Park High School and graduated 6th in her class of 250 students in 2000. Jennifer obtained her Bachelor of Science (BS) degree from the University of Connecticut in May 2003 (majoring in animal science, with a minor in food science). After graduation, she attended the University of Florida on a .33 FTE research assistantship under the supervision of Dr. Keith R. Schneider to obtain her Master of Science (MS) degree in food science. She was awarded her Master of Science in December of 2005. Once graduated, Jennifer plans to work in the food industry specializing in food safety and quality assurance. 96