Salmonella

Salmonella
RESEARCH COMPENDIUM
pork.org • (800) 456-7675

Salmonella Research Compendium • 2
Objective
The National Pork Board continues to serve as a strategic catalyst between U.S. pork producers, researchers and
academics to fulfill its mission of research, education and promotion of pork by doing what’s right for people, pigs, and
the planet. Pork producers have a strong commitment to and long history of producing safe pork.
The objective of this Salmonella Research Compendium is to be a complete guide for producers, veterinarians, packers/
processors, academia and other representatives of the pork and allied industries to reference Pork Checkoff-funded
research focused on this topic. From 1997 to 2018, all Checkoff-funded research pertaining to Salmonella has been
funded by the Pork Safety, Quality, and Human Nutrition Committee, the Swine Health Committee, the Public Health,
Worker Safety Committee, and the Environmental Committee.
Introduction
Worldwide, Salmonella is one of the most common foodborne pathogens. In relation to pork production, Salmonellosis is
the clinical disease in pigs and are asymptomatic carriers. Below (Figure 1), characterizes potential sources of Salmonella
in the pork supply chain (Dickson et al., 2013). With over 2,500 serovars, Salmonella also has potential implications on
human and public health. Salmonella continues to be a challenge for pork producers, packers/processors, retailers,
foodservice and all consumers. Not only does it cause a risk to public health, but has implications on swine health, well-
being, and environmental factors. For these reasons, research in pork safety plays a critical role in supplying a safe,
nutritious, affordable pork product.
Systems perspective on Salmonella sources in pork production and processing.
Manure
Inputs Pork Management
Pigs Production System
Slaughter
Feed/Water Process Trimmings
Transport and and Prevalence Fabrication
Air/Dust
Holding Process Evisceration Process
Insects Farrowing Live Live Process
Rodents Nursery Animal Animal Carcass
Humans Grow/Finish Prevalence Transport Prevalence Prevalence
Holding/Lairage
s
r Pig
Othe en
t
n m
o , )
vir uck en
En (tr ng p
i
ld
ho
Research repeatedly plays a primary role in better understanding of Salmonella this pathogen of interest. Pork Checkoff
is one of the limited sources of funding pork safety research. In total over the past 21 years, Pork Checkoff has funded 97
Salmonella research projects, totaling over $3.7 million dollars.
Additional Salmonella and Pork Safety Resources

• Pork Checkoff Salmonella Fact Sheet
• Pork Checkoff Pork Safety Pre and Post-Harvest Fact Sheets
• Reference https://www.pork.org/food-safety/ for additional pork safety information
• Partnership for Food Safety Education
• American Association of Swine Veterinarians
• American Meat Science Association
Table of Contents
Section 1: Pre-Harvest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-25
• Pathogenesis & isolates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
• Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
• Diagnostic tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
• On-farm prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
• Control strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
• Antibiotic resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
Section 2: Peri-Harvest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26-31
• Associated shedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Section 3: Post-Harvest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32-43
• Prevalence in pork and pork products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32
• Controlling Multi-Drug Resistant (MDR) bacteria in pork . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
• Diagnostic tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
• Risk assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39
Section 4: Environmental Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44-45
All full research projects final reports are available at pork.org/research.
Note: All Pork Checkoff research will be identified by the project number followed by project title. The first two numbers
indicate the year in which the research was funded. The last three or four numbers are for internal record keeping
purposes. For example:
(16-215) Investigation of pathogenicity, competitive fitness, and novel methods for rapid diagnosis of
Salmonella I 4,[5],12:i:- in swine.
From this example shown, the number 16 represents that the National Pork Board Committee funded the research in the
year 2016 and the number 215 represents an internal documentation number.
Section 1: Pre-Harvest
PATHOGENESIS AND ISOLATES
(16-215) Investigation of pathogenicity, competitive fitness, and novel methods for rapid diagnosis of
Salmonella I 4,[5],12:i:- in swine
Since the mid-1990’s, Salmonella Typhimurium has been the predominant serotype of Salmonella enterica identified in
diagnostic samples as a significant cause of clinical disease in swine. However, over the past 5 years, identification of
Salmonella Typhimurium in samples from diseased swine has decreased substantially while identification of the Salmonella
I 4,[5],12:i:- serotype has concurrently increased. The overall goal of this study was to enable the pork industry to better
understand the significance of increasing identification of this emerging serotype as well as provide veterinarians and
producers with improved diagnostic tests to allow for more rapid identification of I 4,[5],12:i:- in diagnostic samples. This
study addressed this overall goal in four ways: 1) direct comparison of the pathogenicity of I 4,[5],12:i:- in swine when
compared to the known pathogenic Typhimurium and lesser pathogenic Derby serotypes; 2) assessment of the fecal
bacterial population over time in Salmonella-infected versus non-infected swine; 3) evaluation of the ability of I 4,[5],12:i:- to
outcompete Typhimurium in swine when co-infected; and 4) development of a rapid PCR test for identification of suspected
pathogenic (i.e.- I 4,[5],12:i:- and Typhimurium) versus lesser pathogenic serotypes of Salmonella in diagnostic samples.
To complete the first two objectives, 72 pigs were inoculated with either Salmonella Typhimurium, Salmonella I 4,[5],12:i:-,
Salmonella Derby or sham-inoculated and followed for up to 28 days following inoculation. During this study, fecal scores
and samples, rectal temperatures, and gross and histopathologic necropsy data were collected to compare pathogenicity
and shedding of Salmonella between serotypes. Sequencing of the bacterial population present in the feces of these
pigs using 16S metagenomic analysis was also completed. The results of this study clearly demonstrate that Salmonella
serotype I 4,[5],12:i:- possesses similar ability as Salmonella Typhimurium to cause significant disease in growing swine
and can be carried in the tonsils and lymph nodes and shed in the feces of infected animals for weeks following exposure
and illness. While Salmonella Derby can also be carried in the tonsils and lymph nodes for extended periods of time,
fecal shedding is limited as is clinical disease. Despite the significant clinical disease caused by infection with Salmonella,
large scale changes in the bacterial populations present in the feces of infected swine identifiable via 16S metagenomics
analysis are minimal. Infection with pathogenic Salmonella serotypes such as Typhimurium and I 4,[5],12:i:- do, however,
lead to a decrease in the overall diversity of the bacterial populations in the feces which could lead to increased
susceptibility to other gastrointestinal disease.
A second animal study utilizing 18 pigs, 12 co-inoculated with both Salmonella Typhimurium and Salmonella I 4,[5],12:i:, and
3 each singly inoculated with either Typhimurium or I 4,[5],12:i:-, was completed to evaluate the comparative competitive
fitness of I 4,[5],12:i:- in the porcine host. When co-inoculated at the same levels, I 4,[5],12:i:- was consistently detected in
the feces of a higher percentage of pigs and at higher levels than Typhimurium. This indicates that serotype I 4,[5],12:i:- may
possess the ability to outcompete serotype Typhimurium when inoculated simultaneously into naïve pigs, which may partly
explain why I 4,[5],12:i:- has been increasingly identified in swine diagnostic samples over the past several years.
Finally, validation of a rapid PCR test for identification of pathogenic Salmonella serotypes was completed utilizing
samples submitted to the Iowa State University Veterinary Diagnostic Lab. Results of this validation shows that PCR
can be used equivalently to culture for Salmonella detection within diagnostic samples. For accurate identification of
Salmonella serotype, enrichment prior to PCR is recommended to ensure that adequate levels of Salmonella are present
in the sample for reliable identification. The use of this PCR in place of standard culture and serotyping, particularly when
combined with enrichment, can reduce the turn-around time for diagnosis of pathogenic serotypes Typhimurium and I
4,[5],12:i:- in swine diagnostic samples from upwards of 6 weeks to a maximum of 2 days.
The combined effect of completion of these objectives will assist producers and practitioners in better understanding
the role of serotypes Typhimurium and I 4,[5],12:i:- in clinical disease in swine. With the use of the newly developed PCR,
veterinarians can more rapidly identify pathogenic Salmonella as the cause of disease within herds to initiate appropriate
interventions to decrease further production losses.
(16-192) From feed to meat: investigation of the prevalence and distribution of Salmonella enterica
serotype I 4,[5], 12:i: - a pathogen of interest in pork
Research project final report still in progress.
Ensuring the safety of pork is essential for producers in order to maintain animal and human health, and also to continue
serving export markets. One barrier to this is the rising occurrence of Salmonella contamination in pork. In order to
best prevent Salmonella in post-harvest pork, the pathogen must first be prevented from entering the farm-to-fork
supply chain. Recently Salmonella enterica serotype I 4,[5], 12:i:- have been linked to swine feed and pork products. The
magnitude of its presence in the U.S. and its pathogenicity are currently unknown. Therefore, the overall objective of this
study was to give to the pork industry a better understanding of the ecology and distribution of Salmonella enterica and
in particular of the serotype I 4,[5], 12:i:-; and collect valuable data for the development of effective intervention strategies
both at pre- and post-harvest level. In order to achieve these overall objectives, the specific deliverables were as follows:
• determine the presence and distribution of Salmonella enterica population and in particular the prevalence of serotype
I 4,[5], 12:i:- in commercial feed mills manufacturing feed;
• characterize the distribution of Salmonella positive isolates in relation to sampling location and establishment-
production associated risk factors.
For objective 1, results indicated that 9 of the total 11 feed mills had at least one Salmonella positive site. From the total isolates,
49 (12.8%) were confirmed Salmonella: two isolates were identified in feed, while the other 47 on equipment and/or surfaces. A
higher pathogen prevalence was observed in fall and summer season. A total of two ST and three STM were identified during
the visits. These isolates showed a similar seasonal presence being found most commonly during fall and summer.
For the second objective mill ID was the significant factor associated with the presence of Salmonella. Differences in
management, geographical location, hygiene practices, quality of incoming raw ingredients, volume of feed produced,
number of workers, and time the facility has been operational were all important variables for pathogen presence. No
differences between mash and pelleted feed were observed, while production flow and plant design were identified as
important in preventing microbial introduction and recontamination of finished feed. Worker shoes also represented an
important vehicle for pathogen spread around the mill: control room and manufacturing floor area (zones with the most
human flow) had the highest percentage of Salmonella samples.
Overall the data gathered in this study shows the potential role of feed and feed mill environment as entry routes for
Salmonella spp, ST and STM into human food chain.
Hygiene, management, production flow, and cross-contamination within facility were all significant factors linked with
pathogen contamination in mills. We found that both the mill and the season were significantly associated with the
presence of Salmonella in the production facilities. These results contribute both to the implementation of biosecurity
plans and other preventative strategies in feed mills and to understand Salmonella behavior at pre-harvest level.
(12-069) Salmonella serovar distribution and persistence in finisher sites

There is a paucity of data explaining the persistence of Salmonella serovars on swine farms. In this study, we evaluated 900
pigs from 18 cohorts of finisher swine over a 3.5 year period in one production company to understand the distribution
and persistence of Salmonella serovars on swine farms and to identify risk factors associated with persistence of
Salmonella serovars in swine. Based on the results of this study, the duration of Salmonella shedding in pigs is associated
with the Salmonella status of the nursery of origin (more positive samples from the nursery, the longer the duration of
shedding), infection at a younger age (pigs first detected as positive at 10 weeks of age shed Salmonella longer than
pigs 12 week of age or older at first positive Salmonella fecal sample) Salmonella serovar (S. Agona was shed for a longer
duration) and the number of treatments in the group of pigs from which the pig originates (above average treatment rate
decreased the duration of Salmonella shedding in swine).
(12-001) Longitudinal study to determine Salmonella serovars and identify risk factors associated with
their dissemination in commercial swine farms
The purpose of this study was to determine what Salmonella serotypes are present in pigs and their environment on
commercial swine farms, and to further characterize and compare Salmonella from different sources. Salmonella isolates
were obtained during a longitudinal study sampling pigs and their environment from 30 commercial swine farms in
North Carolina. This study followed ten cohorts (groups) of pigs from farrowing to slaughter, sampling the pigs, farm
environment, pig carcasses and the slaughter environment at various stages of production. Sampling was carried out from
October 2008 to December 2010 at various stages of production, including once at farrowing (7-10 day old), twice at each
of nursery (4 and 7 weeks of age) and finishing stages (16 and 26 weeks of age), and finally once at slaughter. During the
farrowing stage, a cohort of 35 healthy piglets per farm (4 piglets/sow) were selected and ear tagged for identification;
subsequently, sampling followed the same cohort of pigs at different sampling stages during farm and slaughter
stages. Salmonella isolates collected during this study were sent to the National Veterinary Services Laboratories (NVSL)
to determine the serotype. A representative subset (n=272) of Salmonella was genotyped using Pulsed-Field Gel
Electrophoresis (PFGE) to determine the specific strain, or PFGE fingerprint profile, at different stages of production.
Comparisons were then made between PFGE strains and serotypes found in pigs and the farm environment. Even
though the focus of this grant was at the farm level, for providing more in depth information to the scientific community
we compared fingerprint profiles of Salmonella isolated from carcasses of the same pigs at slaughter with the farm
isolates. Serotyping revealed 22 different serotypes found in pig and environmental samples on farm and at slaughter,
with Salmonella Typhimurium being the most common. Genotyping analysis revealed 47 clusters containing 100% similar
Salmonella isolates among pig and environmental isolates, including feed, water, soil, lagoon, floor swabs and slaughter
lairage both within and between cohorts. In addition, 41 unique strains were also detected. This indicates that certain
serotypes and strains are present in pigs and their environment throughout the pork chain, irrespective of the farm or
stage of production. We also found evidence highlighting the clear role played by the environment in the persistence and
dissemination of Salmonella to conventionally reared pigs at farm and slaughter.
(09-120) Molecular Basis of Salmonella Competition in Broth Culture

Salmonella remains an extremely important bacterial pathogen to the swine industry. It is a significant pathogen
affecting swine health and also represents one of the most important foodborne pathogens affecting people. A key
aspect of Salmonella control is the use of cultivation methods in the laboratory. Samples must be cultured to isolate
specific Salmonella strains, and this step is necessary to understand the spread of Salmonella throughout the swine
production system. Unfortunately, conclusions about Salmonella transmission are highly dependent on the performance
characteristics of these cultivation methods, and recently, we quantified a disturbing fact: the probability of detecting a
specific Salmonella strain in a sample might have very little to do with its concentration in the sample but more to do with
its ability to compete in the cultivation media and with the specific mixture of Salmonella strains present in the sample.
The overall objective of this project was to characterize the bias that cultivation media has on Salmonella detection and
enumeration. To accomplish this, we conducted a series of competition experiments using Salmonella serovars and strains
from the swine production system. We had the following specific aims:
1. Establish whether four Salmonella strains isolated from the swine production system exhibit heterogeneity in growth
and competitive fitness during cultivation in broth media.
2. Determine which genes are either up or down regulated in these Salmonella strains in the presence of different
cultivation broth media and during competition with other Salmonella strains.
To conduct our experiments, we used 4 different Salmonella enterica serovars originally isolated from swine: S. Agona, S.
Derby, S. Mbandaka, and S. Typhimurium. When grown individually in different media that are routinely used to culture
Salmonella from swine samples, we found the following. In one broth, S. Derby exhibited the fastest growth and therefore
appeared to have the potential to outcompete the other strains. However, in a second broth at 37°C, S. Derby did not
grow at all. In this broth at 37°C, S. Typhimurium grew the fastest. In a third medium at 37°C, S. Agona grew the fastest. S.
Derby was able to grow slightly in the early hours of the growth curve. When the serovars were grown in a Most Probable
Number format, there were again major differences among strains. In general, the serovars had a more difficult time
growing at 42°C than at 37°C, even though 42°C is often used. S. Derby again did not grow at all. These results began to
demonstrate that important Salmonella serovars could be entirely missed using standard protocols.
When the strains were competed against each other in 2, 3 or 4-way competitions, there was a clear ordering of
competitive ability among the strains, but this order depended on the media and temperature. Once again, S. Derby did
not grow well and was never detected in any competition.
Finally, we identified specific genes in these strains that might be contributing to the differential growth characteristics.
An understanding of the genetic basis of these differences could help us design a more appropriate cultivation protocol
(with more appropriate media) to accurately culture all the Salmonella strains that might be present in swine samples.
This study confirmed what we had previously found in a pilot study: that the probability of detecting a specific Salmonella
strain in a sample might have very little to do with its concentration in the sample but more to do with its ability to
compete in the cultivation media and with the specific mixture of Salmonella strains present in the sample. We will now
proceed to identify optimal strategies for cultivating Salmonella from swine samples.
(08-034) Identifying genes associated with Salmonella shedding to increase pork safety through
improved genetic resistance- year 2 of 07-062
The long-term goals of this project are to use genomics tools to identify genetic variants that are associated with
decreased Salmonella shedding on farms; and to use these variants to select for animals that shed fewer bacteria and
so are less likely to cause abattoir carcass contamination. Thus, the aim is to directly improve pork safety, as well as
decrease other on-farm contamination by contaminated fecal matter that may be used as fertilizer. To achieve the long
term goals, we worked on 4 specific objectives: 1) generate a large bank of Salmonella fecal shedding data and matched
DNA samples using US industry pigs; 2) create DNA tests for genetic variation at candidate genes, selected based on our
previous data on porcine gut lymph node and whole blood gene expression patterns in response to Salmonella infection;
3) perform genotyping using new and established DNA tests on US industry Salmonella-positive–control groups; and 4)
test associations of genotypes with Salmonella incidence and integrate data from epidemiology and genomics projects to
understand Salmonella contamination.
For objective 1, during a 2 year period a total of 462 belly flap tissue samples have been collected that include 163 pigs
that shed Salmonella in their feces 7 days before marketing as well as 299 farm-mate negative controls matched for
farm-visit cohort where a positive sample was found. Shedding of Salmonella was determined by qualitative fecal testing.
Objective 1 was performed in collaboration with a large epidemiology study funded by the USDA-NRI Competitive
Grants Program. We emphasize here the extraordinary contribution of this USDA-funded project; it would not have
been possible to collect such a large number of Salmonella positive field samples within the current funding structure
for standard NPB projects. In our laboratory, the DNA from Salmonella-positive and negative-control pigs belly flap
tissue have been isolated and was used for genotyping. We also used other 3 porcine populations that we created using
prior NPB funding or current funding from the USDA-NRI Functional Genomics program, such as the NADC-40 pig and
NADC-77 pig Salmonella Typhimurium challenge populations; and the 228 pig IAH-Compton Salmonella Choleraesuis
challenge population kindly provided by the Pig Improvement Company. In objective II we selected candidate genes for
analysis of genetic variants (single nucleotide polymorphisms, SNPs). As candidates we selected genes that, based on our
previous research, are involved in the porcine response to Salmonella infection. For objective III during the 1st-year period
we genotyped 29-31 SNPs across the 4 populations and identified 18 SNPs with minor allele frequency (MAF) of 15% or
higher in at least 2 populations. During the 2nd year period, we have genotyped a set of 31 additional SNPs across our
pig populations and identified 10 new genetic variants with MAF of 15% or larger in at least 2 populations. In all, during
the two year period we selected and genotyped across the 4 populations 62 SNPs in 53 genes that are involved in porcine
response to Salmonella infection and confirmed 28 SNPs segregating in at least 2 populations. This output exceeded
the original proposed goals of 10 segregating SNPs per year of funding. To genotype this larger number of SNPs, we
adopted a Sequenom technology due to its increased throughput and reduced cost. In objective IV, we have newly
applied a statistical approach to measuring Salmonella shedding/tissue colonization levels over the 3-week period after
infection. Statistical analyses revealed several SNPs associated with Salmonella fecal shedding or tissue colonization in
pigs. Fecal shedding associated SNPs were CCT7 #3 (p=0.041), GNG3 (p=0.029), PGD (p=0.047) and HP #2 (p=0.0002) in
the field population, AMT (p=0.005) in the NADC-40 pigs population and PGD (p=0.001) in the NADC-77 pigs population.
In the IAH-Compton population three SNPs in the ACP2 gene were associated with Salmonella burden in spleen (ACP2#1,
p=0.013, ACP2#2, p=0.026, ACP2#3, p=0.033); and in the NADC-40 pig population, a SNP in the EMP1 (p=0.002) gene
was associated with bacterial load in ileo-cecal lymph node. The genotyping and statistical analysis data identified genetic
markers that are potentially useful for selecting animals that shed fewer bacteria and are less likely to cause pen-mate
Salmonella contamination in a farm and on a slaughter plant. Additional validation of these SNPs should be performed,
but this project has provided novel information for the pig industry to implement in strategies for selecting pigs with
reduced shedding and/or disease susceptibility.
(07-062) Identifying genes associated with Salmonella shedding to increase pork safety through
improved genetic resistance
The long-term goals of this project are to use genomics tools to identify genetic variants associated with decreased
Salmonella shedding on farms; and to use these variants to select for animals that shed fewer bacteria and so are less
likely to cause abattoir carcass contamination and thus improve pork safety. To achieve these long term goals, we worked
on 4 specific objectives: 1) generate a large bank of Salmonella fecal shedding data and matched DNA samples using
US industry pigs; 2) create DNA tests for genetic variation at candidate genes, selected based on our previous data on
gut lymph node gene expression patterns in response to Salmonella infection; 3) using new and established DNA tests,
perform genotyping on U.S. industry Salmonella-positive–control groups and laboratory challenge populations; 4) test
associations of genotypes with Salmonella incidence and integrate data from epidemiology and genomics projects to
understand Salmonella contamination.
For objective 1, during a 15-month period a total of 178 belly flap pig tissue samples from 33 farms have been collected
that includes 59 Salmonella positive and 119 negative samples as determined by qualitative fecal testing. Objective 1 is
an ongoing project performed in collaboration with a large epidemiology study funded by USDA-NRI. In our laboratory,
DNA from Salmonella-positive and negative-control pig tissue is being isolated as soon as samples are collected. In
objective 2 we confirmed 11 new genetic variants (single nucleotide polymorphisms, SNPs). To analyze genetic variants
in the objective 3 work, three porcine populations with Salmonella shedding data were used: 1) the field pig population
(objective 1); 2) a 40 pig NADC Salmonella Typhimurium challenge population; and 3) a 228 pig Compton Salmonella
Choleraesuis challenge population. In addition to confirmed SNPs we also selected to be genotyped 38 new putative
SNPs in 25 genes involved in regulation of porcine response to Salmonella. To genotype this larger number of SNPs, we
adopted Sequenom technology due to its increased throughput and reduced cost. For the first round of genotyping, a
total of 29 selected SNPs have been combined and will be genotyped using all 3 porcine populations, a total of 378 DNA
samples. The objective 4 statistical association analysis will be performed upon completion of this genotyping project.
The genotyping and statistical analysis data will be used to generate genetic markers useful in selecting animals that shed
fewer bacteria and are less likely to cause pen-mate Salmonella contamination in a farm and on a slaughter plant, thus
providing novel information for the pig industry to implement in strategies for selecting pigs with reduced shedding and/
or disease susceptibility.
(05-176) Identifying the Genetic Profile of the Salmonella-carrier Pig to Improve Food Safety and
Decrease Pre-Harvest Disease
Pigs that shed Salmonella in their feces are an animal health issue, a food safety problem and an environmental
contamination risk. Our research program investigates the undesirable carrier state of Salmonella in pigs in order to
identify approaches to control both clinical and sub-clinical (carrier) infections. Our research has identified pig genes
that respond to infection with Salmonella. A DNA sequence variation was identified in one of the Salmonella-response
genes (CCT7). When we compared the sequence of the CCT7 gene with the Salmonella shedding status of 40 infected
pigs, we found that pigs with a specific DNA sequence were less likely to shed Salmonella in their feces than pigs with a
different DNA sequence. Thus, pigs with this specific DNA sequence may be more resistant to infection with Salmonella.
Furthermore, the amount of a specific immune compound (interferon-?) in the blood of the infected pigs also correlated
the Salmonella shedding status of the pigs as well as specific blood cells that fight infection. Identifying factors in the pig
that control the ability of the animal to combat disease will uncover markers for classifying potential carrier pigs as well as
targets for disease diagnosis, intervention and prevention.
EPIDEMIOLOGY
(11-159) Does the inclusion of distillers dried grains with solubles (DDGS) in the diet of grow-finish pigs
affect their susceptibility to and colonization with Salmonella enterica?
As an alternative to counteract the increased feed costs, dried distillers grains with solubles (DDGS) have been
increasingly included in pig diets. Much research has been conducted recently to evaluate growth performance and
carcass characteristics associated with feeding DDGS to pigs. However, little is known about the potential effect of
DDGS on the susceptibility to infection or colonization with pathogens. Therefore, two experiments were conducted to
determine if inclusion of DDGS in the diet of grow-finish pigs affects their susceptibility to or the intestinal levels and
shedding of Salmonella. In experiment 1, 36 pigs (12 pigs/treatment) were assigned to 3 treatments: Control diet with
no corn DDGS, diet with 20% corn DDGS, or diet with 40% corn DDGS. After an adaptation period of 2 weeks, each
pig was inoculated with Salmonella and euthanized after 6 hours to determine their susceptibility to the challenge. In
experiment 2, 40 pigs (20 pigs/treatment) were assigned to 2 treatments: Control diet with no corn DDGS or diet with
30% corn DDGS. After 2 weeks, each pig was inoculated with Salmonella; individual fecal samples were collected during
5 weeks, and pigs were euthanized at 3 and 5 weeks post-challenge to determine intestinal colonization. In experiment
1, no differences among treatments were observed on the susceptibility to Salmonella infection. In experiment 2, most
pigs shed Salmonella at one of the fecal samplings during the study period, with control pigs having a higher cumulative
shedding frequency than pigs receiving the diet with 30% DDGS. There was no difference between treatments regarding
the average Salmonella fecal shedding level. Also, no difference between treatments was found on the frequency or levels
of Salmonella in intestinal samples collected at 3 or 5 weeks post-challenge. In conclusion, dietary inclusion of corn DDGS
does not alter the susceptibility to or colonization with Salmonella of grow-finishing pigs.
(03-132) Bioinformatics Based Genome Comparison of Six Salmonella spp. to Provide a Foundation for
the Development of Reliable Molecular Epidemiological Methods.
Salmonella research is the top priority for the pig industry especially related to pre-harvest reduction of pathogens
with potential public health significance. The pre-harvest food safety research for Salmonella proposed here falls
under the category of control strategy. The research proposed here is highly relevant and applied in nature. Without
the groundwork, foundation and benefit of comprehensive genetic bioinformatics based analyses; many shotgun
approaches based upon methods that are applicable to other microorganisms for instance, will ultimately fail when
used for Salmonella spp. This project makes available to the industry and food safety scientists a powerful comparative
sequence and genome comparison databases that can be used to develop epidemiological methods that are specific to
Salmonella spp. rather than attempting to adapt methods that have proven to be of use with other organisms. Salmonella
spp. because of how closely related the serovars tend to be presents a significant challenge to the development and
applicability of molecular epidemiological methods. This project represents the primary and quite simply the only
logical foundation step that will ever result in the development of a method that fulfills all the requirements of a superior
molecular epidemiological method for Salmonella spp.
(98-194) Colonization Pattern of Salmonella Typhimurium DT104

Salmonella Typhimurium definitive phage type (DT) 104 has a unique antimicrobial resistance pattern (R-type) with
multiple resistance observed for ampicillin (A), chloramphenicol (C), streptomycin (S), sulfonamides (Su), and tetracycline
(T) (ACSSuT). Recovery from humans and animals has increased over the past several years and has become the
predominant phage type in a number of regions world-wide. Although dissemination appears to be widespread,
factors influencing this occurrence have not been defined. We challenged pigs with either a wild-type penta-resistant
S. Typhimurium DT104 (Group R) or a pan-sensitive S. Typhimurium DT104 (Group S) by commingling challenged pigs
(n=2/group) with sentinel pigs (n=13/group) for each group. Control pigs (Group C) were not exposed to Salmonella.
Pigs were monitored for signs of clinical illness and rectal shedding of Salmonella was also monitored. Pigs from each
group were necropsied at weeks 1, 2, and 3 post-challenge. Tonsil, liver, cecum and cecal contents, ileocolic lymph
nodes (ICLN) and ileocolic junction (ICJ) were collected aseptically for qualitative bacteriology. Cecal contents were also
evaluated quantitatively. Rectal shedding of Salmonella in sentinel pigs was sporadic in Group S (average 9% positive for
days 0 thru 6). For Group R, rectal shedding increased over time (27% and 100% positive on days 1 and 6 respectively)
in sentinel pigs exposed to the resistant strain and averaged 32% positive over the period evaluated (days 0 through 6).
Levels of Salmonella being shed into the environment were determined for challenged pigs and sentinel pigs by group
on day 3. Feces (n=1) from Group S had 1.4 log10 CFU/g Salmonella while feces (n=2) from Group R pigs contained 1.7
log10 CFU/g. Salmonella levels were too low to quantify in fecal composites from sentinel pigs in either Group R or S on
day 3. The percent Salmonella positive tissues recovered from Group S was 50, 40 and 8% for necropsy at weeks 1, 2 and
3, respectively. The percentage Salmonella positive tissues recovered from Group R was 76, 64 and 56% for necropsy at
weeks 1, 2 and 3, respectively. Over the course of the study 35 and 65% of the tissues were positive for Group S and R,
respectively. Levels of Salmonella in the cecal contents declined over time in Group S from 2.2 log10 CFU/g at week 1 to
undetectable levels for weeks 2 and 3. Levels of Salmonella in the cecal contents persisted in Group R and were 1.8, 1.5
and 1.7 log10 CFU/g Salmonella for weeks 1, 2 and 3, respectively. This suggests that acquisition of multiple resistance
has altered the colonization potential of the strain and may explain the dissemination of the organism among many
different species. Further studies are needed to determine the factors affecting colonization of Salmonella with different
antimicrobial resistance patterns in order to develop strategies for altering colonization.
(98-146) Antibody to Salmonella spp., culture status, and associated risk factors among slaughtered pigs
Antibodies to Salmonella in pigs can be detected, indicating that pigs were exposed to one or more serotypes of
Salmonella. We measured antibodies to Salmonella spp. in 3,354 pigs from 145 commercial pig herds in the upper
Midwest. A mixed-antigen ELISA test was applied to meat-juice samples, detecting antibody response to many Salmonella
species. We classified pigs into relatively lower (category I) and higher (category II) antibody response, with the threshold
set by a cluster analysis. Overall, 45.6% of pigs were in category II (95% CI 43.9-47.3%). The variation in within-herd
prevalence was large (SD 30.6%), suggesting a diversity of Salmonella cycling patterns among farms. Prevalence of
category II antibody response was a poor predictor of Salmonella culture prevalence at slaughter. Herd or barn factors
associated with an increased risk of high antibody category II prevalence were partial slotted or solid flooring (OR 3.75,
p=0.09), poor growth performance (OR 6.8, p = 0.08) and bird access to the barn (OR 4.48, p=0.05). Batch pig flow (OR .35,
p=0.09) and access to rodents (OR 0.2, p = 0.07) were associated with a decreased risk of antibody category II status. These
findings suggest that improved hygiene may reduce Salmonella cycling and/or exposure on farms. The apparent protective
association with rodent exposure is unexpected and not biologically plausible. It is possible that farms with rodent
exposure tended to have other, unmeasured characteristics that could account for the apparent protective association.
Further study of these factors is needed to test their effectiveness at reducing Salmonella exposure and cycling.
(97-1877) Effects of methods on the isolation of Salmonella from swine feces and implications for design
and interpretation of epidemiologic studies
We conducted a series of experiments on fecal samples collected from commercial swine farms to evaluate the effects of
several methodologic factors on detection of the Salmonella. For practical purposes, we constructed these studies around
the routine procedure used for Salmonella isolates in our laboratory. Factors examined were: fecal sample weight, fecal
sample storage (no storage, 4°C, -15°C), and delayed secondary enrichment were. In addition we compared the standard
method (Method 1) used in our laboratory [10 g fecal/buffered peptone water pre-enrichment/selective enrichment in
Rapport Vassilidis (RV) broth] with another method (Method 2) used by ourselves and other in the USA (1g sample/primary
enrichment in tetrathionate and Hajna GN broth/ secondary enrichment in RV broth). This included a comparative study
using matched samples processed in our laboratory and in the laboratory of Dr. Paula Cray (ARS, Athens Georgia).
The weight of feces sampled had a marked linear effect on the detection of Salmonella. Estimated relative sensitivities
ranges from 8.7% for rectal swabs to 78.3% for 25g fecal samples. Studies based on rectal swabs are likely to yield very
unreliable results. Immediate processing of samples yield the best recovery of Salmonella, although storage at 4°C for
6 days did not significantly reduce detection. Immediate processing of samples is prudent where possible, but storage
of feces for a few days at 4°C will not markedly affect results. Freezing of fecal samples resulted in significant reduction
of detection and should be avoided when possible. The use of delayed secondary enrichment yielded variable results
among trials, but generally increased the detection of Salmonella. Direct comparison of Method 1 and Method 2
indicated comparable results, with Method 1 generally yielding higher recovery of Salmonella. However, when conducted
on samples of equal weight, Method 2 had significantly better detection than Method 1. The choice of methods can
markedly affect the results of fecal sample culture. The preferred methodology for epidemiological studies will be
determined by many factors including logistics and cost. Our data highlights the imperfect sensitivity of culture methods,
and the need for researchers to consider the sensitivity of their bacteriologic methods in the design of field studies based
on fecal culture. The collective results from these studies provide a useful framework for comparing recently published
data of Salmonella prevalence in the US, and for investigators designing field studies of Salmonella in swine.
DIAGNOSTIC TOOLS
(09-094) Development and validation of molecular-based tools to differentiate attenuated Salmonella

choleraesuis vaccine strains from field isolates
The objective of this research was to develop molecular diagnostic methods for differentiating live, attenuated Salmonella
Choleraesuis vaccines from virulent Salmonella-like bacteria. Samples submitted to the Minnesota Veterinary Diagnostic
Lab (MVDL) as part of routine diagnostic workup were used for test development. 115 suspected vaccine isolates and 101
known Salmonella Choleraesuis isolates were tested using the newly developed test. 22 additional Salmonella serotypes
and 14 unrelated pathogens were also tested to demonstrate the specificity. The multiplex test for the Argus live
vaccine strain of Salmonella Choleraesuis was successful in quickly and specifically identifying the Argus vaccine and in
differentiating it from virulent strains of Salmonella Choleraesuis. This test is currently being offered for routine diagnostics
at the Minnesota Veterinary Diagnostic Laboratory.
(06-185) Salmonella sample cryopreservation enables epidemiologic study, quantification of bacteria

and further characterization of bacterial isolates
Salmonella control requires practical means to reliably test for shedding of the organism. Fecal samples offer the potential
to sample on farms or at slaughter plants. However, methods to culture Salmonella are time consuming and require
immediate processing. This means that samples must either be processed at the site, or transported immediately to an
appropriate laboratory. The objective of this study was to develop reliable methods to freeze samples so that the logistics
of sampling could be improved in future studies or other applications. We inoculated feces from growing pigs with strains
of Salmonella previous derived from healthy pigs at slaughter. Freezing samples decreased the number of Salmonella
recovered using conventional microbiologic technique. Addition of glycerol reliably increased the recovery, that is,
decreased the loss of culturable Salmonella due to freezing. The addition of sterile skim milk, Buffered Peptone Water
and LB Broth also enhanced recovery, although not as consistently as glycerol. Extending samples with 20% glycerol was
reliably helpful; 40% glycerol was detrimental in some experiments. Adding a water based diluent (Phosphate Buffered
Saline) before freezing reduced Salmonella recovery; this effect was lessened if glycerin was also added. The findings of
this project suggest that for certain applications where financial or practical constraints make storage of samples before
testing desirable, freezing with the addition of 20% glycerol and 20% skim milk should increase recovery of Salmonella.
(05-063) Sequencing of 10 potential molecular epidemiological targets from a collection of food safety
relevant Salmonella spp.
The development and testing of molecular epidemiological tools for pre-harvest food safety applications is an important
and rapidly evolving issue. This means that it is vital to track bacterial contamination of our food to its source, whether that
contamination originated during production, transport, holding/lairage or even within processing plants, delivery, stores,
or at home. As an example Salmonella contamination found in meat products could be tracked back to a failed water
supply on a farm if a successful molecular tracking method could be developed. By identifying sources we can prevent
further contamination of our food originating from that source. As part of this project our goal was to take steps toward
the development of improved molecular epidemiology tools using new whole genome informatics. Specifically our goals
were: 1) To sequence ten (10) identified heterogeneous target regions (at least 3x coverage) from 40 Salmonella enterica
serotypes 2) Develop specific PCR primers that will amplify these target regions in all serotypes.3) Test the targets for
their ability to sensitively and specifically identify or distinguish isolates of S. enterica from our culture collection. We
successfully completed all of these objectives and exceeded the goals by including an additional 10 serotypes in the
sequencing objective. We have developed a set of primers that are able to amplify target regions in all of the serotypes
we have tested. We have also sequenced a set of test isolates that have allowed evaluation of a computer database
genetic homology comparison method using the data generated in this study. Our results show that we can identify
the S. enterica serotypes test isolates correctly, which is a valuable tool in itself. With the increase cost of serotyping a
genotyping method would be extremely valuable to the industry. We are now using a neural network approach to develop
a model that will allow evaluation of complicated set of data that is generated during the homology comparison step
in this process. Thus, we have developed a method that shows great potential as a genotyping method for S. enterica
serovars and has potential as a method for source tracking if it becomes possible to continue to expand and improve the
genetic databases and models we have initiated.
(05-053) Determination of the diagnostic sensitivity and specificity of a reference method for
Salmonella spp. detection in swine fecal samples of known infection status
Salmonella enterica is a bacterium that causes disease in many animals, including swine and human beings. This organism
can be transmitted from swine to human beings through direct contact, or through indirect contact such as by eating
pork products that have been contaminated with the organism (i.e., foodborne infections). Many European countries have
implemented surveillance methods to detect pigs that are shedding Salmonella species as they go to slaughter, and these
standards may eventually be adopted by the United States. In addition, diagnostic methods to detect Salmonella species
are used to determine cause of illness in sick pigs. The traditional way to detect Salmonella species is to perform bacterial
culture in a laboratory. This method, while traditional, widely used and somewhat standardized, can consist of many
variations. If the sample probably contains very few bacteria, there can be a pre-enrichment step, where all organisms are
allowed to grow to high numbers, followed by a selective enrichment step where the Salmonella are allowed to grow but
other types of bacteria are inhibited. Finally, the samples are put onto a solid medium that contains biochemicals which
help to identify Salmonella. This is especially useful when the sample is fecal material or intestinal tract, which contains
many different types of bacteria.
It has been shown that variations in the steps used, or the type of media used for each of these steps in the culture
procedure can greatly affect the amount and types of Salmonella recovered. It has also been shown that the same
protocol, used in different laboratories, can result in differences in recovery of Salmonella. This project was conducted
in order to evaluate five different culture protocols, used in one laboratory, on samples collected from pigs of known
infection status (positive or negative). This research project generated information that can then be used to compare
additional culture methods, besides of demonstrating how different culture methods perform for the recovery of
Salmonella from naturally contaminated swine fecal samples.
(03-048) Development of a rapid detection method for Salmonella serotypes using a flow cytometric assay
The use of serological assays to monitor the health status of swine herds has been used for several decades. The primary
assays employed have included agglutination, neutralization, and/or ELISA assays to detect the presence of serum
antibodies to a specific infectious agent. The current work indicates that a flow cytometric assay can be used to evaluate
serum samples for antibodies reactive with multiple antigen in a single reaction tube. The ability to detect antibody
responses to five separate Salmonella endotoxin preparations in a single tube was shown. There are several advantages of
this type of assay over traditional plate ELISA assays. For example, different set of beads coated with a variety of antigens
can be mixed in a single reaction tube. The beads could be coated with bacterial, viral, parasitic, or fungal antigens and
combined in a single reaction tube. In addition, the amount of serum sample required for the assay is small (less than 100
µL) regardless of the number of antigen-coated bead sets used in the assay. Lastly, the use of this technology should be
useful diagnostically as well as in monitoring the health status of livestock animals.
(01-107) A Rapid and Specific Test for Salmonella Serovars with Particular Reference to Salmonella
choleraesuis serovar C1
A polymerase chain reaction based enzyme linked immunosorbent assay (PCR-ELISA) was developed to identify
Salmonella serovars A, B, C1, C2, and D. Primers were selected from the rfb gene cluster, which is responsible for
biosynthesis of O antigens of Salmonella lipopolysaccharide. Salmonella isolates (n=203) were tested using the PCR-ELISA
procedure. Of those Salmonella isolates, 156 isolates had been serogrouped previously. These isolates were used to
determine the sensitivity and specificity of this PCR-ELISA procedure. DNA from all isolates was amplified using the PCR
procedure for selected serovars and amplified products were visualized on agarose gels, as well as subjected to the ELISA
procedure. The sensitivity of this procedure to correctly identify Salmonella serovars was 92% and the specificity was
99%. Eighty-eight percent of serovar D/A, 91% of serovar B, 89% of serovar C1, and 100% of serovar C2 were identified
correctly with this procedure. Of the remaining 47 isolates that were not serogrouped, 5 were identified as serovar B, 7
as serovar C1, 1 isolate as serovar D/A and 34 isolates were neither B, C1, C2 or D/A. Results of this study indicate that
the PCR-ELISA procedure is a rapid and accurate method for serogrouping Salmonella isolates. Utilization of the PCR-
ELISA procedure for Salmonella serogrouping would aide the swine industry in identification, surveillance, prevention and
control of Salmonella on the farm.
(01-063) Validation of a Diagnostic Test to Identify Salmonella enterica in Swine: Preliminary

Comparison of Fecal Culture and Serum Antibody Detection
The goal of this study was to describe and compare serum antibody response (as determined by the Danish Mix-ELISA
(DME) and the SalAD) to fecal shedding of S. enterica (SE) during the growing period in commercially raised, naturally
infected swine.
We accomplished this through: 1) longitudinal investigations of 5 groups of individually identified growing pigs in two,
three-site swine production systems and 2) crosssectional sampling of 48 groups of market pigs within one-month of
slaughter. Thirty different serotypes were isolated from the pigs sampled in this study; the most frequently isolated
serotypes were Typhimurium and Typhimurium var Copenhagen. Pig prevalence estimates by fecal culture ranged from
0-72.9% for fecal culture, 0-84% for the Danish Mix ELISA and 8-100% for the SalAD ELISA.
Visual analysis of graphical data of the longitudinal investigation demonstrates that prevalence estimates based on serum
antibody detection methods and fecal culture were similar in pattern throughout the growing phase, and that higher
cut-off values more closely reflected fecal prevalence. There was poor agreement between fecal culture and the serum
antibody detection methods for individual pigs (kappa values ranging from 0.017-0.093, and sensitivity and specificity
ranging from 0.207 to 0.592 and 0.487 to 0.896 respectively). There was a small correlation between seroprevalence
estimated by DME and fecal prevalence estimates, and no evidence of correlation between fecal prevalence and
seroprevalence estimated by the SalAD in the cross-sectional comparisons of prevalence near the time of slaughter.
(00-063) Growth and Immunomodulatory Effects of the Seaweed Ascophyllum Nodosum in Salmonella-
Infected Pigs
A series of in vitro studies was conducted to evaluate the effect of A. nodosum extract (ANE) on activation of pig
splenocytes and alveolar macrophages. In the first set of studies, splenocytes and macrophages were incubated with
concentrations of ANE from .1-100 μg/mL. Concentrations of TNFa and PGE2 were evaluated as measures of macrophage
activation and interleukin 10 (IL-10) and interferon gamma (INFg) were evaluated as measures of splenic lymphocyte
activation. As positive controls, macrophages were stimulated with bacterial lipopolysaccharide (LPS) and splenocytes
were incubated with the mitogen concanavalin A (Con A). ANE failed to stimulate secretion of PGE2 or any cytokine by
cultured immune cells. In a subsequent set of experiments, concentrations of ANE as high as 10 mg/mL were evaluated.
Again, ANE did not alter IL-10 production by splenocytes. However, PGE2 was increased (P < .05) relative to control wells
in macrophages treated with 10 mg/mL ANE after 3 and 24 hours in culture. The results of our in vitro studies demonstrate
conclusively that ANE, when presented directly to immune cells across a wide range of concentrations, generally failed
to activate macrophages and lymphocytes. These immune cells are key players in both cell-mediated and antibody-
mediated immune function. Very high concentrations of ANE enhanced production of the inflammatory mediator PGE2,
but never to the degree achieved by bacterial LPS. An in vivo study also was conducted with dietary ANE. A total of 95
pigs (initially 15 lb and 17 d of age) was used in a 28 d growth experiment to determine the effects of ANE on weanling
pig growth performance and immune function in response to enteric disease challenge with Salmonella Typhimurium (ST).
Experimental treatments were arranged in a 2 x 4 factorial with main effects of disease challenge (control vs. ST challenge)
and dietary addition of ANE (0, 0.5, 1.0, and 2.0% of diet). Results suggest little beneficial effect of dietary ANE on growth
performance or immune response in the presence or absence of ST challenge.
(99-230) Reduction of Salmonella by Bacteriophage Treatment

It is known that Salmonella can disseminate into the body organs of pigs within a few hours post infection. This rapid
dissemination of Salmonella in pigs has become a recent concern because it may increase Salmonella contamination
of pork products. Salmonella infection in pigs prior to slaughter may be required by direct contact with Salmonella
derived from the environment or from other infected pigs. Therefore, any intervention strategy that reduces or prevents
the rapid dissemination of Salmonella in pigs prior to slaughter may be crucial to reduce the prevalence of food-borne
salmonellosis in humans via the ingestion of Salmonella contaminated pork products. In order to address to use of phage
as an intervention strategy against Salmonella, we conducted phage treatment in experimentally infected pigs by using
a classical procedure of phage therapy. This procedure involved a cocktail of Salmonella-specific phages from sewage
samples of swine farms and administrated the phage cocktail to pigs. The phage cocktail consisted of 26 isolated phages.
The efficacy of phage treatment was related to the route of administration and time elapsed after phage treatment.
However, these data showed that this particular phage cocktail did not significantly reduce the level of Salmonella in
pigs. Based on the preliminary data, we used a single Salmonella-specific lytic phage (Felix 0-1) as a possible phage
candidate for intervention strategy against Salmonella. For the experiment, three week-old pigs received 2x1010 plaque
forming units of Felix 0-1 phage both orally and intramuscularly at three hours after challenge with S. Typhimurium in the
concentration of 5x108 colony forming units. Nine hours post challenge, blood, liver, lung, spleen, ileocecal lymph node,
tonsil, and cecum content samples were taken from sacrificed pigs. The numbers of S. Typhimurium in the samples were
quantitatively found by direct culture method. Salmonella Felix 0-1 phage treatment significantly reduced the amount of
S. Typhimurium in tonsil and cecum contents as compared to control. Thus, Salmonella-specific phage treatment can be 2
considered as a short-term intervention strategy to reduce the rapid dissemination of Salmonella in swine.
(99-136) A Rapid, Specific Test for Salmonella Subtypes

A polymerase chain reaction based enzyme linked immunosorbent assay (PCR-ELISA) was developed to identify
Salmonella serovars A, B, C1, C2, and D. Primers were selected from the rfb gene cluster, which is responsible for
biosynthesis of the O antigens of Salmonella lipopolysaccharide. Forty-eight isolates obtained from porcine feces or
lymph nodes following challenge with Salmonella enterica serovar Typhimurium or Salmonella choleraesuis were tested
using the PCR-ELISA procedure. DNA from all isolates were amplified using the PCR procedure for selected serovars
and amplified products were visualized on agarose gel electrophoresis, as well as subjected to the ELISA procedure.
Those isolates that were identified as positive with the PCR-ELISA with Salmonella primers for serogroup B had a mean
absorbance reading of 2.14 +/- 0.57. Negative controls and non-Salmonella bacteria had a mean absorbance reading
of 0.27 +/- 0.15. Those isolates that were identified as positive by the PCR-ELISA assay with Salmonella primers for
serogroup C1 had a mean absorbance reading of 2.75 +/-0.59. Negative controls and non-Salmonella bacteria had a
mean absorbance reading of 0.86 +/- 0.32. Results of the ELISA procedure were verified by agarose gel electrophoresis.
All isolates were identified by biochemical and phenotypic characteristics. Of the 48 isolates evaluated, 36 isolates were
serovar B, 2 isolates were serovar C1, and 10 were neither serovar A, B, C1, C2 or D. Results of this study indicate this PCR-
ELISA procedure appears to be a rapid and accurate method for serogrouping Salmonella isolates.
ON-FARM PREVALENCE
(16-120) Estimating the National prevalence of Salmonella spp. In lymph nodes from market hogs and sows
(16-074) Determination of the Salmonella enterica prevalence in lymph nodes from market hogs and sows
The goal of this study was to determine the prevalence of Salmonella enterica in lymph nodes in sows and market hogs.
Paired superficial inguinal lymph nodes were collected from twenty one commercial facilities in the northern and southern
regions of the United States. Salmonella prevalence was detected in 12.55% of the porcine lymph nodes tested. A subset
of samples was typed using whole genome sequencing. The distribution pattern of the serotypes identified through
whole genome sequencing indicated that multiple serotypes are prevalent in the lymph nodes of swine. Additionally,
these isolates had antimicrobial genes that provided resistance to 4 classes of antibiotics. One of the isolates was
multidrug resistant with resistance against three antibiotic classes. These results indicate that pork and related food
products could act as a potential risk for salmonellosis to the public.
(16-059) National Survey of Salmonella Prevalence, Serotypes, and Antimicrobial Resistance in Porcine
Lymph Nodes
The primary objective of this study was to benchmark the national prevalence rates of Salmonella in the lymph nodes
(LNs) of sows and market hogs. LNs were collected across the United States from twenty-one commercial facilities,
categorized by geographical regions as north or south. Twenty-five paired superficial inguinal lymph nodes were collected
at each facility and shipped to South Dakota State University. Salmonella prevalence rates in the northern region were
37.0 and 6.4% for sows and market hogs, respectively. In the southern region, 4.8% of sow samples and 13.0% of market
hog samples were Salmonella positive. Data on types of chilling methods used at each facility also were collected. In
the northern region, prevalence rates of Salmonella across chilling types were distributed as follows: 20.0, 2.7, and 1.3%
positive samples for conventional, other, and blast chill methods, respectively. Additionally, in the south, there were 20.0%
positive samples for conventional, 12.0% for other chill methods, 0.0% for blast.
(15-070) An analysis of primary research and meta-analysis to investigate temporal changes in

serotypes of Salmonella in US-based pigs and pork products
The goal of our project was to determine if there are important changes in the prevalence of Salmonella serotypes in
swine. This is important because if the serotypes remain unchanged on 20 years, then we know that there are no changes
required in our currently effective Salmonella control approaches. However, new serotypes are emerging or patterns
are shifting, this might signal a change in the ecology of Salmonella and we would need to determine if the controls we
are remain effectives. The pork industry does currently do not have evidence that Salmonella control programs need to
be serotypes specific, however it is important to be proactive and understand which pathogens are circulating in swine
populations and found of pork products.
We used 4 longitudinal datasets which looked at Salmonella from 1996 to 2014. We observed decreasing proportions of
S. enterica serovar Typhimurium, serovar Derby, and serovar Heidelberg and increasing proportions of S. enterica serovar
4,[5],12:i:-, serovar Infantis, and serovar Johannesburg in swine. We also observed positive correlations for the yearly
changes in S. enterica serovar 4,[5],12:i:-, serovar Anatum, and serovar Johannesburg between swine and human data; in
S. enterica Worthington between avian and human data; and in S. enterica serovar 4,[5],12:i:- between bovine and human
data were observed.
These data suggested that some serovars are emerging while others are decreasing. In particular, serovar 4,[5],12:i:- is
found commonly in swine submitted to diagnostic laboratories in the past 5 years. As this organism has been associated
with large outbreaks of disease, this is an important finding. These data also suggested that veterinary diagnostic
laboratory data are potentially timely for detecting changes in serotypes and that serovar 4,[5],12:i:- is common in swine
submitted to VDLs.
(14-288) Assessment of International Swine Programs for On-farm Salmonella Control

The recognition of pork as one of the many sources of Salmonella infections has resulted in a variety of government
and industry initiated science-based programs to minimize its occurrence in pigs, pork meat, and pork products.
This bacterium can inhabit swine intestines, lymph nodes, and be shed in their feces. Once it resides in pig herds,
contamination of the slaughter environment, and eventually the meat can occur. Depending upon the specific nation, pre-
or post-harvest procedures were implemented to reduce Salmonella prevalence at specific stages of the farm-to-table
continuum. There have been considerable assessments to determine those interventions that are the most efficacious
and cost-effective Salmonella controls, and where to best apply them to achieve a measurable public health benefit.
Exponent, Inc. (Exponent) was retained by the National Pork Board (NPB) to critically review the effectiveness and assess
the costs of such efforts.
(04-201) Determination of Salmonella Classification Levels for Selected Midwestern Swine Herds
Utilizing Abattoir-based Samples and Comparison of Danish Herd Classification Stability Overtime
Since 1995, the Danish swine industry has conducted an on-farm Salmonella control program to categorize production
sites as to their risks for Salmonella contamination. This classification has been used to schedule transport and to harvest
herds of similar statuses to reduce antemortem Salmonella cross contamination. Danish herd classifications for Levels
I, II, and III have remained relatively static at approximately 95%, 3.3%, and 1.6%, respectively (Nielsen et al., 2001). In
2002 similar values (91%, 8%, and 1%) were obtained from a large Midwestern USA sample. In 2004 a similar sampling
scheme compared 502 herds from 2002 and 2004 and found classifications for Levels I, II, and III to be 79%, 14, and
6%, respectively. This increase was also reflected in an approximately 2X increase in total samples testing positive for
Salmonella antibodies. Analysis of these data suggests that the prevalence of Salmonella within production units was not
temporally stable and that the prevalence of Salmonella exposure in Midwestern swine differed between these surveys
for small volume but not large producers. The results highlight that classifications assigned to producers in 2002 were
not likely to be the same in 2004. These findings have implications for Salmonella control schemes that use on-farm
classification schemes to impose marketing restriction on herds based upon antibody prevalence levels.
(04-108) Serological Analysis of Toxoplasma gondii, Trichinella spiralis and Salmonella enterica among
niche market antimicrobial-free and conventionally reared pigs
Three globally important zoonotic foodborne pathogens including Salmonella, Trichinella and Toxoplasma, were studied.
Serological analysis of these pathogens was conducted on samples collected from two distinct pig production systems:
antimicrobial-free, niche market type (ABF) and conventional (indoor-reared). The results from this preliminary study enabled
us to understand that all the three pathogens were more commonly present in pigs that were reared in antimicrobial-free,
out-door, niche-market type of environment than the commercial conventional indoor types of herds. This was found to be
consistent in all the three states that were included in the current study: North Carolina, Ohio and Wisconsin.
(04-100) Salmonella prevalence during the finishing production stage

This research project was conducted with the objective of determining if the Salmonella prevalence at the end of the
finishing production stage correctly represents the Salmonella status of a lot, as a high or low Salmonella prevalence lot.
An important discrepancy between bacteriologic and serologic prevalence estimates was found. Although all lots of pigs
studies were persistently negative through bacteriologic testing, they were all serologically positive at slaughter. How
can all the studied lots be bacteriologically negative for Salmonella during the entire growing-finishing production stage
(i.e., over a period of 17 weeks), but be serologically positive at slaughter? This question leads to re-thinking the complex
ecology and epidemiology of Salmonella in swine herds, and a consequent formulation of fundamental questions: How
well does serology represent the real microbial food safety risk at the time of slaughter? How long do infected pigs carry
and shed Salmonella? How long does the serologic response to a Salmonella infection last in pigs? Although the number
of published studies on the epidemiology of Salmonella in subclinically infected pig herds has tremendously increased
over the past decade, longitudinal studies following both the bacteriologic and serologic statuses of pigs are still
necessary to understand their dynamic and inter-relationship. Moreover, to design effective pre-harvest monitoring and
control programs, an understanding of the transmission of Salmonella, taking into account a number of variables (e.g.,
serotype, age of infection, infectious dose, route of infection, diet configuration, environmental exposures, etc.) on farms,
is essential. This research generated valuable information, besides of raising critical questions that need to be answered in
order to advance our understanding of the ecology and epidemiology of Salmonella in swine populations.
(03-026) Role of gilts in the introduction and transmission of Salmonella in swine production systems
Three swine production units in Illinois were examined with respect to the role of purchased gilts in introducing new
genetic variants of Salmonella into a herd, and the risk of incoming gilts acquiring Salmonella infection from resident pigs.
On each farm a cohort of at least 100 gilts was followed from day of arrival until the time of introduction into the breeding
section of the herd. Fecal and floor samples were obtained from the gilts and from resident pigs in adjacent pens or
rooms after each movement of the cohort gilts to a new location, and cultured for Salmonella. Polymerase chain reaction
(PCR) techniques were used to detect and genetically classify Salmonella isolates to identify transmission of the pathogen.
Salmonella was detected at varying times and locations in the production systems examined, although only a small
percentage of pigs (usually less than 10%) were shedding Salmonella at any time. The close genetic similarity of isolates
obtained from samples collected within a room or building on a given farm visit indicates that there was limited mixing
of Salmonella from different sources on the farm, that is, there is usually a low level of transmission among pigs within
different locations on the farm. Close genetic relatedness of some isolates from gilt samples with some isolates from
resident pig samples suggests that a small but detectable amount of transmission of Salmonella was occurring between
gilts and resident pigs that resided in adjacent locations.
Whereas most Salmonella infection is subclinical and therefore rarely a problem in reducing productivity, the proliferation
of virulent strains of Salmonella, some of which may have antibiotic resistance, is a serious concern for swine and human
health. Separation of pigs from different sources through all stages of production, even in the breeding herd, as well as
biosecurity measures restricting movement of personnel across sections of the herd with pigs from different sources,
coupled with exclusive use of artificial insemination, is recommended as the optimal strategy to minimize transmission of
different genetic variants of Salmonella across different sections of the herd.
(02-187) Determination of Salmonella Classification Levels for Selected Iowa Swine Herds and an
Evaluation of Production Risk Factors for Classification Status
This study represents the first attempt to classify typical Midwestern production sites for Salmonella spp. sero-prevalence
using the Danish classification system. With 1,131 farms represented, this data provides a useful snapshot of Midwestern
Salmonella prevalence and can be used as a benchmark for future efforts. The data suggest that Midwestern herds are
similar in their distribution with respect to Salmonella sero-prevalence classifications as Danish herds. It also indicates
that a significant percentage of farms present a small number of market swine for harvest that have experienced prior
Salmonella exposure.
This study compared information obtained from production practice questionnaires generated by producer self-
administration and by interview. Interview-based applications gave better repeatability and content validity. The
main modifications of the current questionnaires should include: (1) clarification of terminology by using interviews
as a method of questionnaire delivery and (2) removal of highly correlated questions. When modifying a production
practices questionnaire, efforts should be made to clearly capture farm structure differences between on-site and off-site
production systems, and how they affect biosecurity responses. Another area for improvement in future questionnaires is
the use of common terminology.
The most serious limitation of any questionnaire has not been addressed by this study, and that is the true accuracy or
criteria validation. No gold standard exists to compare the responses with on-farm Salmonella exposure, and therefore,
the responses from either the self-administered or the interview questionnaire may not accurately reflect true on-farm
Salmonella risks.
(00-100) Evaluation of Methods to Differentiate Herd Responses to Salmonella

An evaluation of serology as a method to determine the Salmonella exposure levels of finisher age pigs utilized two major
designs. First, individual serum samples and rectal swab cultures, and pen fecal samples were collected from pigs and
pens in three herds known to have different levels of Salmonella. These specimens were used to compare rectal swab
and pooled pen fecal culture methods as well as four serologic tests; the Danish mix-ELISA (DME), the Maxivet Diakit
ELISA, the Biotest Salmotype ELISA, and the SalAD (Salmonella Antibody Detection) ELISA. None of the tests were
sensitive enough to be used to identify Salmonella exposure in individual pigs. If applied on a herd basis, pooled pen
fecal cultures and the DME, Diakit and Salmotype serologic tests were found to be acceptable for categorizing herds as
to level of Salmonella exposure. Those three serologic assays were evaluated with their established calibrated methods
while only raw OD readings were available for the SalAD analysis. SalAD results may be different with regression line
calibration with controls or by using techniques such as S/P ratios. In the second experimental method, finishing pigs in
64 herds in 12 U.S. states and Canada were monitored serologically with the DME for a period of 10 months. The overall
Salmonella seroprevalence of Canadian herds was significantly higher than herds in the Western, Central, and Northern
U.S. Seroprevalence was significantly higher in the fall and winter samplings compared to the spring samplings, and fall
significantly higher than summer seroprevalence levels.
(98-148) Salmonella Abatement in the Pork Chain

The project focused on the Salmonella infection-contamination-infection cycle on swine farms. The results reveal that
the Salmonella occurrence on swine farms is much more dynamic (prevalence and serotype patterns changing over time)
than previously thought. Up to 18 different serotypes can be found on a single farm. The “exchange” of Salmonella spp.
between infected animals and their contaminated environment plays a key role in determining farm-specific Salmonella
patterns, which can and must be identified before implementing intervention measures that are capable of reducing
the amount of Salmonella carried into the food chain via slaughter hogs. Subtyping of selected serotypes has shown
that serotyping alone does not allow to conclude on infection chains. DNA-fingerprinting methods help to identify
infection sources and to develop targeted prevention measures. Comparing the resistance patterns of Salmonella strains
isolated from slaughter animals to strains that were isolated from farm environment samples has shown that antimicrobial
resistance in Salmonella is wide-spread and not only limited to strains from animals that are potentially treated with
antimicrobials.
CONTROL STRATEGIES
(16-113) Salmonella DIVA vaccine for cross-protection against Salmonella serovar I 4,[5],12:i:-:
minimizing antibiotic usage and protecting swine and public health
(11-082) Analysis of Salmonella Pre-Harvest Control Strategies for Domestic & International Markets
The purpose of this project was to provide a comprehensive description of the efficacy of mandated control programs for
pre-harvest Salmonella spp. used by major pork exporting/importing nations. Such information enables the industry to be
aware of the approaches used that may affect trade and food safety.
The 1st and 2nd aspects of the project were to describe mandated Salmonella spp. specific pre-harvest control
approaches in major pork exporting/importing countries. We searched grey and peer reviewed literature for reported
programs and contacted individuals in these countries. The information was difficult to find and verify. Our conclusion is
that no major pork producing countries have mandatory Salmonella spp. specific pre-harvest control programs except
Denmark. Denmark requires mandatory declaration of Salmonella spp. status prior to sale and testing of purchased feeds
for Salmonella.
The 3rd aspect of the project was to collate estimates of the prevalence of Salmonella spp. pre-harvest in the United
States. We used data from two prior reviews and an updated review of the literature. We collated estimates of the
prevalence of Salmonella positive herds and the within herd prevalence ( Figure 1. )
The 4th aspect of the project was to summarize the efficacy of mandated programs. The Danes suggests their combined
pre-harvest and post-harvest approach is effective but the importance of the pre-harvest component is not clear1-4.
An assessment of the Belgium voluntary program which focused unilaterally on primary production concluded “The
evaluation of this Salmonella action plan demonstrated that it had only little effect on the level of Salmonella infection
in pigs.”5 The European Food Safety Commission has published two cost benefit assessments of pre-harvest control of
Salmonella spp. in finishers6 and breedings7 and concluded “On the basis of current scientific advice and the experience
of Member States, it is not possible at this time to demonstrate cost-beneficial interventions to reduce Salmonella
infections at EU level in either breeding pigs or slaughter pigs, or in combinations of both herds. Sensitivity analyses
indicate that positive cost-benefits can be found only in extreme scenarios.
The 5th aspect of the project was to summarize the findings about the effect of pre-harvest interventions and assess the
GRADE system for making recommendations. We used published reviews of Salmonella pre-harvest interventions and
the GRADE process to assess the efficacy8. The GRADE process uses 4 factors to reach a conclusion for adoption of an
intervention: the quality of the evidence, the cost benefit, the balance of benefits and harms and the balance of values
and preferences. The panel concluded it was unlikely that any intervention assessed would be strongly recommended to
swine production entities for adoption based either on lack of evidence for an effect, or very low cost versus benefit (i.e.
would cost much more than the benefit).
(09-099) Use of cationic peptides as feed additives to improve innate immunity and reduce gut
colonization with Salmonella and Campylobacter in weaned pigs
The objectives of this study were to determine if the addition of a cationic peptide to the feed formulation of weaned
pigs decreases Salmonella and Campylobacter colonization of the pig’s gut and to investigate the effects of using cationic
peptides as feed additives in weaned pig rations on the innate immune response of weaned pigs. Pigs were fed a feed
ration containing 24 ppm of the BT TAMU cationic peptide for the duration of each part of the studies. In the Salmonella
studies, pigs in both control and peptide groups were given Salmonella orally. Daily rectal swabs were taken and pigs were
euthanized and cultured for the presence of Salmonella on day 7 after infection. Pigs fed the peptide had reduced fecal
shedding of Salmonella and reduced Salmonella in the tissues compared to control pigs that did not receive the peptide.
In addition, pigs fed the peptide had higher daily weight gains than did pigs that were not fed the peptide. No significant
reductions of Campylobacter were found in pigs fed the peptide with preexisting Campylobacter infections. In innate
immunity studies, one mechanism swine leukocytes use to kill microbes, called an oxidative burst, was found to be higher
in pigs fed the peptide versus pigs that did not receive the peptide on all days it was measured. Leukocyte degranulation,
another mechanism leukocytes use to kill microbes, was not found to be different between control and peptide-fed pigs. In
summary, the addition of the BT TAMU peptide to weaned pig feed resulted in reduced Salmonella shedding in feces and
reductions in tissue colonization. The peptide also stimulated the immune system of pigs fed the peptide, as indicated by
an increased oxidative burst response in these pigs. Although no effects against Campylobacter were observed, this may
have been due to the fact that all pigs used in the studies had preexisting Campylobacter infections. If Campylobacter-
free pigs were available, the results may have differed. These studies indicate the BT TAMU peptide may be an effective
alternative, immunologically-based strategy to the use of antibiotics to reduce Salmonella in swine.
(06-003) Toll-like receptor agonist administration for the prevention of Salmonella colonization of the
swine gut
The main objectives of this research project were to determine if the oral administration of certain novel innate (“natural”)
immune enhancers would protect piglets against Salmonella infections and if the oral administration of these immune
enhancers had beneficial effects on some important immune functions in young piglets. One immune stimulant, CpG
DNA (CpG), is a component of bacterial DNA that is recognized by the pig’s immune system as a non-self threat and
causes the pig to mount an immune response to the CpG DNA. By mounting a response to the CpG DNA, the pig’s
immune system is also then primed to take on other pathogenic microbes that it may encounter, such as Salmonella. The
other immune stimulant, flagellin (Flag), is a protein produced by bacteria and is a component of the bacteria’s flagella.
Flagellin acts in much the same way as CpG on the pig’s immune system. Each of these innate immune stimulants are
recognized by the pig’s immune system through members of a highly evolutionarily conserved family of receptors called
the toll-like receptors (TLR). TLRs are now recognized as being highly important not only for initial encounters with
most microbes, but are also important for the host to mount an acquired or “memory” response to pathogens during
subsequent encounters.
(05-193) Use of naturally-occurring bacteriophage to reduce Salmonella in swine prior to harvest

Salmonella is one of the most common causes of foodborne illness in humans and can also impact swine production
efficiency. Bacteriophage are viruses found in nature that can kill bacteria, including Salmonella. Generic phages were
found to be fairly widespread in commercial swine, but only 6 out of 360 commercial swine were positive for anti-
Salmonella bacteriophage. When these phage were used in pigs that were artificially infected with Salmonella, the
Salmonella populations in phage-treated pigs were lower compared to untreated controls. Additionally, fewer pigs were
colonized by Salmonella in the cecum in the phage-treated groups. While we do not suggest that we have a complete
solution for Salmonella, our data indicates that the concept of using phage to reduce Salmonella in swine is valid and
feasible. We are continuing to examine commercial swine to obtain more potent anti-Salmonella phage to be able to
maximize this potential pathogen reduction strategy. We gratefully thank the National Pork Board and our U.S. pork
producers for their generous financial assistance in exploring this important area of improving food safety.
(04-030) Control of Salmonella by segregated early weaning, topical disinfection and administration of
sodium chlorate
The results of this study document that sodium chlorate / nitrate / lactate, used alone or combination with early weaning
and/or topical disinfection can be effective in reducing Salmonella concentration in feces and cecal content and in
reducing Salmonella prevalence in feces and ileocecal lymph nodes. Since the compound has low toxicity and can be
applied to pigs of differing ages, these findings suggest that chlorate should be considered as a preharvest intervention
to reduce Salmonella shedding in swine. The treatment was ineffective in breaking the cycle of transmission at weaning,
even when done in combination with other treatments. The Salmonella reduction was prolonged through the 9 days of
follow-up testing. This may allow for increased flexibility in the application of chlorate as a pre-harvest intervention. In
sum, topical disinfection and CHLORATE may be useful tools for farms that practice segregated weaning, and where sow
to piglet transfer of Salmonella is an important source of infection in nursery pigs. However, this would be effective as a
food safety intervention if and only if the duration of protection is sustained through the growing period, a question that
was not addressed in the current study.
(02-046) Corn Particle Size and Pelleting Influence on Fecal Shedding and Enteric Colonization of
Salmonella Typhimurium
Grinding corn to a finer particle size has been shown to have beneficial effects on growth, feed efficiency, and digestibility.
The trial demonstrated improvements in growth and efficiency. However, in contrast to previous research, pelleting did
not result in an improvement in growth and feed efficiency. Using this model, we were unable to detect influences of feed
processing on fecal shedding and colonization of mesenteric lymph nodes with Salmonella. Therefore, it appears that the
increased risk of finer grinding and pelleting of feeds associated with Salmonella shedding reported in other studies may
be due to factors other than those confined to the intestinal tract environment.
(01-077) Lactobacillus acidophilus as Probiotic to Control Salmonella in Swine

The purpose of this project was to determine if a selected probiotic culture of Lactobacillus acidophilus would exert
control of Salmonella in swine. Several cultures of L. acidophilus previously isolated from swine (5 different animals) were
tested for the ability to inhibit the growth of Salmonella Typhimurium in a laboratory medium. One of the cultures of L.
acidophilus (strain L-23) which exerted 90% inhibition was chosen for use in the feeding trials.
For the feeding trials, immediately after being weaned, pigs were assigned to two groups. Pigs in group 1 were fed
individually (daily) 10 ml portions of sterile milk and those in group 2 were fed in a like manner milk containing cells of L.
acidophilusL-23 for 10 days. The pigs in both groups also were fed a starter ration without antibiotics. On the tenth day,
the animals in both groups were challenged orally with a culture of Salmonella Typhimurium (1x109 cells per animal). All
pigs continued to receive the appropriate milk with or without the L. acidophilus daily. Fecal samples were obtained from
each animal immediately before and on days, 2, 7, 14, 21, and 28 after being challenged with the pathogen. Ten grams
of each fecal sample was analyzed for numbers of Salmonella. The trial was repeated three times using cultures of S.
Typhimurium obtained from three separate research intuitions where each had been shown to cause Salmonella infections
in young pigs. Unfortunately, none of the three cultures of S. Typhimurium caused infections in any of the pigs in our three
trials. Because of this it was not possible to determine whether or not the probiotics culture could control this pathogen.
While the feeding trials did not provide an answer concerning the effectiveness of the probiotics, the results do suggest
that a culture of Salmonella causing infection in pigs from one swine herd will not necessarily do so for pigs in a different
herd. In future trials we plan to use both oral and nasal route of infection with the Salmonella. This has been suggested as
a means to enhance the effect of the pathogen.
(00-103) Applied studies on immune stimulation by Lactobacilli for the reduction of human foodborne
Salmonella and replacement of subtherapeutic antimicrobials
In the present studies, two Lactobacillus isolates were separately used to grow yogurt cultures. These yogurt cultures were
fed on a daily basis to four-week-old pigs. On the eighth day of yogurt feeding, the pigs were inoculated with Salmonella
enterica subsp. enterica serovar Typhimurium. Yogurt feeding continued for either 14 or 27 days, during which time
Salmonella levels were monitored by culture. In two of six trials, feeding live Lactobacillus paracasei subsp. paracasei GS-1
yogurt cultures significantly reduced levels of Salmonella in tonsil and fecal samples from swine.
ANTIBIOTIC RESISTANCE
(15-072) Effects of antibiotics on intestinal and extra-intestinal multidrug-resistant and pan-susceptible

Salmonella in swine
Antibiotic resistance is a major health concern in both human and veterinary medicine. Antibiotics already used to
maintain healthy swine may also play a role in reducing Salmonella; however, the relationship between antibiotics and
antibiotic resistant bacterial populations needs to be further understood. The overall objective of this study was to
investigate the effects of antibiotics on the intestinal and extra-intestinal bacterial populations of swine challenged –
both orally and transdermally – with pan-susceptible and multidrug-resistant (MDR) Salmonella. A Salmonella challenge
experiment was conducted on 32 pigs conducted in two replicates. In each replicate four pigs were assigned to each
of the four treatment groups; chlortetracycline treatment, ceftiofur treatment, ceftiofur and chlortetracycline treatment,
and an untreated control group. Pigs were challenged on days 1 and 3 with a 104 culture of pan-susceptible and MDR
Salmonella enterica serovar Senftenberg transdermally and 104 culture of pan-susceptible and MDR Salmonella enterica
serovar Derby orally. A single dose of ceftiofur was given on day 5 and chlortetracycline was given in feed from day 5 to
day 18 for the maximum treatment period of 14 days. Intra-rectal fecal samples were taken daily throughout the study
and on day 19 the pigs were euthanized and tonsils, mandibular lymph nodes, peripheral lymph nodes (superficial front-
and hind-limbs), intestines, and gut-associated lymph nodes were collected. Samples were cultured for Salmonella using
standard enrichment and selective techniques. Salmonella isolates were confirmed and further characterization included
serogrouping and PCR of antimicrobial resistance genes specific to the challenge strains. 16S rRNA metagenomics
analyses of fecal samples from days 1, 5, 12, and 19 were conducted to observe differences in the bacterial communities.
The overall fecal prevalence of Salmonella between treatment groups varied significantly for non-enriched samples
(p<0.05) but did not vary significantly for enriched samples (p=0.61). The overall Salmonella prevalence in lymph nodes
was 8.1% for non-enriched samples and 20.3% for enriched samples and prevalence in lymph nodes using enrichment
methods did not vary significantly (p=0.14) between treatment groups. All 658 fecal isolates were serogroup B which
is indicative of Salmonella Derby. We found 98 serogroup B lymph node isolates and 13 serogroup E4 (indicative of
Salmonella Senftenberg) lymph node isolates. The majority of the antimicrobial resistance gene PCR results matched
well with the expected serogroup and we found that 49.7% of the fecal isolates harbored qnrB. Among the lymph
node serogroup B isolates, 69.4% harbored qnrB and 5 of the 14 serogroup E4 isolates harbored blaSHV. The greatest
difference in microbial communities was found between the control and the ceftiofur/clortetracycline treatment groups
and between days 5 and 19, though the differences were not significant. Overall we found that Salmonella infection
can quickly and efficiently create persistent enteric infections with constant or intermittent shedding in a controlled
environment, and that the intradermal route does not appears to be an efficient route of infection for this bacteria.
However, even in this scenario, Salmonella was recovered from peripheral lymph nodes after 14 days of antibiotic
treatment. We found that antibiotic treatment reduced the number of Salmonella in feces below detection limits but
did not eliminate Salmonella shedding. Further analyses are needed to investigate differences in the isolates recovered
between treatment groups to explore differences in the pan-susceptible and multi-drug resistant Salmonella populations.
(07-076) Enhanced virulence and treatment of multiple antibiotic resistant Salmonella choleraesuis in swine
Previous studies revealed that protozoa (amoebae) can augment the virulence (ability to cause disease) of certain
Salmonella strains that are resistant to multiple antibiotics. This phenomenon has not yet been observed in swine
because many of these strains have low virulence in domestic pigs. However, resistance to multiple antibiotics was
recently discovered in Salmonella choleraesuis, a strain that causes the most severe type of disease in swine. Enhancing
virulence in Salmonella choleraesuis, would be devastating for the swine industry and this recent discovery is the basis
for this project. The first objective of this project was to evaluate the possibility that protozoa can augment the virulence
of multiple antibiotic resistant Salmonella choleraesuis. The second objective was to determine the best antibiotic for
treating this infection. The research involved evaluating the ability of protozoa to enhance Salmonella choleraesuis in
the laboratory and in swine. Laboratory studies involved measuring the ability of Salmonella choleraesuis to penetrate
host cells, i.e. cellular penetration is an important part of Salmonella virulence, after exposure to protozoa. Swine studies
involved orally infecting 10 day-old pigs with Salmonella choleraesuis exposed to protozoa. Swine were monitored for
signs of disease and necropsies were performed in order to determine the amount of Salmonella in the animals. Some
pigs were treated with either of two antibiotics: ceftiofur or amikacin. Results from these studies revealed that protozoa
are capable of enhancing the cellular penetration of Salmonella choleraesuis by approximately 700%. Animal studies
revealed that protozoa-exposed Salmonella choleraesuis were capable of causing disease at 24 hours earlier compared
to pigs infected with Salmonella choleraesuis that had not been exposed to protozoa. Tissue samples revealed that
Salmonella choleraesuis was ten times more prevalent in swine infected with Salmonella choleraesuis exposed to
protozoa. Both ceftiofur and amikacin ameliorated signs of disease (fever, diarrhea, and lethargy) although ceftiofur-
treated pigs had a smaller load of Salmonella choleraesuis. The results indicate that multiple antibiotic Salmonella
choleraesuis can be more virulent after exposure to protozoa. Protozoa are water-borne common microbes and thus
it appears that the combination of protozoa and Salmonella choleraesuis can lead to a dramatic course of Salmonella
infection in swine. Ceftiofur seems to be the most appropriate treatment for this infection.
(02-141) Risk Factors for the Occurrence of drug resistant Salmonella spp. in commercial swine production
Resistance to two antimicrobial drugs frequently used in swine production, tetracycline and sulfa drugs, was common
among Salmonella isolates from commercial farms. Since these isolates were from apparently healthy pigs, this resistance
may not be important for therapy of disease in pigs. However, administering these drugs could result in increased
prevalence of these strains, and potentially increase the number of Salmonella in slaughtered pigs. The presence of
antimicrobial resistant strains probably increases the risk of passing antimicrobial resistant strains down the food chain,
especially if the farms administer antibiotics to which these strains are resistant. Very few isolates, less than one percent,
were resistant to antimicrobials more important for human therapy, including ciprofloxacin, amikacin, ceftriaxone,
and trimethoprim/sulphamethoxazole. Thus, the human infections by these strains should be responsive to therapy,
should foodborne infection occur. The presence of one gene accounted for the majority of tetracycline resistance.
However, this gene was apparently linked to resistance genes or factors for a number of antimicrobials. Therefore, the
use of tetracycline has the potential to select for Salmonella strains resistant to multiple antimicrobials. Farmers could,
therefore, inadvertently select for multi-resistant strains when using only one antimicrobial, such as tetracycline. This
could potentially compromise therapy on farms and pose an increased risk to pork food safety. Increased mixing of pigs
appears to increase the risk of resistant strains, since multiple nurseries flowing to a single grower/finisher increased the
risk of finding resistant Salmonella strains. The biological reasons for this cannot be determined by this research method,
but bears further research. One possible cause is simply from mixing of pigs with more diverse Salmonella strains from
separate populations. Other causal factors, not recorded by the survey, may also be associated with these farm types.
Increased barn size, but not increased farm size, was associated with increased risk of antimicrobial resistant Salmonella.
The presence of larger numbers of animals in a single barn may result in an increased variety of strains of bacteria at
placement, increasing the likelihood of starting with more resistant strains. In addition, larger barns may be managed in
different ways from smaller barns. These associated management factors could cause increased spread of antimicrobial
resistant strains. Feeding pelleted feed was associated with antimicrobial resistance. Pelleted feed has been also
associated elsewhere with increased risk of Salmonella shedding. Coupled with the findings here, additional research on
the mechanisms by which pelleted diets might cause (or be associated with) these increased risks is warranted.
(02-037) The Role of Sub-Therapeutic Chlortetracycline on the Prevalence and Antimicrobial Resistance
of Salmonella Enterica Isolated from Swine
There is increasing scrutiny regarding the use of antimicrobials in animals as a result of increasing antimicrobial resistance
in human pathogens, and in particular foodborne outbreaks caused by multiple resistant foodborne pathogens (for
example Salmonella Typhimurium DT104). The perceived risk of transfer of resistance from bacteria colonizing animals to
pathogens of humans has led to the ban of sub-therapeutic antimicrobials in animal feeds in many European countries
and has prompted suggestions for a moratorium on use in the US. The overall objective of this research was to evaluate,
in a controlled manner, the effect of sub-therapeutic chlortetracycline on antimicrobial resistance and pig performance
on commercial swine farms. In this study, there was no effect of sub-therapeutic chlortetracycline on the prevalence or
antimicrobial resistance of Salmonella isolated from swine. There was a small, but statistically significant increase in the
proportion of fecal flora isolates resistant to tetracycline, ampicillin and ceftriaxone in the fecal flora of swine that received
sub-therapeutic chlortetracycline in their diets, but no effect was seen in resistance to gentamycin. The public health
implication of this change in resistance is uncertain and unanswerable by this project. Further efforts to measure the risk
attributable to antimicrobial use on farms to antimicrobial resistant infections in humans is necessary.
(01-033) Effect of Antibiotic Regimens on Resistance in Salmonella and Other Bacteria in Swine
To compare the effects of antimicrobial regimens on resistance of Salmonella Typhimurium and non-pathogenic bacteria
associated with swine, 29-day-old weaned pigs, were challenged with Salmonella Typhimurium and divided into 6
groups and housed in identical SEW nursery rooms. Pigs were assigned by room to treatments including a control (no
antibiotic); oxytetracycline in the feed for 30 days, increasing gradient application of oxytetracycline, pulse dosing with
oxytetracycline, rotation with oxytetracycline, apramycin sulfate, and sodium sulfamethazine in the drinking water and
carbadox in the feed, and rotation with oxytetracycline, apramycin, and neomycin sulfate in the drinking water, and
neomycin sulfate in the drinking water. Fecal samples were collected prior to antibiotic treatment and throughout the
grow/finish period for culture of the Salmonella challenge organism and E. coli. Isolates were tested for resistance to test
antibiotics using minimum inhibitory concentration analysis. DNA based techniques, including PCR and PFGE were used
to detect resistance genes and to determine relatedness of isolates. Antibiotic regimen affected resistance of E. coli but
not of the Salmonella challenge organism. Molecular analysis indicated that the apramycin resistance gene is associated
with bacterial plasmids, whereas tetracycline resistance genes are associated with the chromosome in E. coli.
(99-114) Antimicrobial use and antimicrobial resistance of Salmonella isolated from nursery
and finishing pigs
Salmonella is an important foodborne pathogen that can exhibit resistance to multiple antimicrobial agents. We
evaluated antimicrobial resistance of 1314 Salmonella isolates obtained from 47 groups of pigs raised in multiple-site
production systems in two different companies. We found most of the isolates (85%) to be resistant to tetracycline,
with less frequent resistance to other antimicrobials such as ampicillin (47%), amoxicillin/clavulanic acid (23%), and
chloramphenicol (22%). We further found that specific Salmonella serovars commonly exhibited certain single and
multiple resistance patterns. For example, over three-quarters of chloramphenicol resistant isolates were of the serovar
Typhimurium var. Copenhagen, and most isolates of this serovar (54%) were resistant to the combination of ampicillin,
amoxicillin/clavulanic acid, chloramphenicol, pipericillin, and tetracycline. These patterns were found in isolates obtained
from several farms of both companies and at different farm visits. This suggests that resistance patterns in Salmonella
obtained from swine depend upon the serovar rather than the site of origin.
(98-222) On-farm antimicrobial use and patterns of antimicrobial resistance of Salmonella isolates
collected on farms and at slaughter plants
Salmonella is an important foodborne pathogen that can exhibit resistance to multiple antimicrobial agents. We
evaluated antimicrobial resistance in 370 Salmonella isolates obtained from pigs at 5 farms and 486 isolates from these
pigs after commercial slaughter. Samples of feces from defined groups of finishing pigs on commercial farms, and
cecal and mesenteric lymph node samples from the same groups of pigs after slaughter were cultured for Salmonella
and tested for resistance to a panel of 11 antimicrobials. We found that the prevalence of antimicrobial resistance
varied among serotypes, the most striking example being the occurrence of multiple resistance in 94% of isolates of S.
Typhimurium, compared with 11% of isolates of all other serotypes. Concordance between Salmonella serotypes from
on-farm samples and those serotypes isolated after slaughter varied widely among farms, suggesting that risk of exposure
to Salmonella during transport and lairage remains a concern under contemporary industry conditions. Slaughter plant
studies based on bacterial culture do not provide a reliable index of the Salmonella status of commercial swine farms nor
their associated patterns of antimicrobial resistance.
(98-218) Characterization of Multiple-Antimicrobial Resistance among Swine Salmonella

Forty-two Salmonella isolates obtained from diseased swine were genetically characterized for the presence of specific
antimicrobial resistance mechanisms. Twenty of these isolates were S. Typhimurium DT104 and the other 20 were non-
DT104 Salmonella strains. PFGE was initially employed to determine the genetic relatedness among the Salmonella
isolates and revealed twenty distinct genetic patterns among the 42 isolates. However, all DT104 isolates fell within two
distinct genetic clusters. Other Salmonella isolates genetically grouped together within their specific species.
All DT104 isolates displayed the pentamer resistance phenotype to ampicillin, chloramphenicol, streptomycin,
sulfamethoxazole, and tetracycline (ACSSuT). Resistance to sulfamethoxazole, tetracycline, streptomycin, kanamycin,
and ampicillin was most common among the non-DT104 Salmonella isolates. The presence of integrons among these
Salmonella isolates was also investigated. All DT104 strains contained two chromosomal integrons of 1000 and 1200
base pairs. DNA sequencing revealed that the two integrons contained genes encoding resistance to streptomycin and
ampicillin, respectively. None of the non-DT104 strains showed this same pattern although several strains possessed
integrons of 1000 bp or larger. Several of the non-DT104 strains did not possess any integrons. This research suggest
that integrons contribute to antimicrobial resistance among swine Salmonella but are not responsible for all resistance
phenotypes observed.
Section 2: Peri-Harvest
ASSOCIATED SHEDDING
(07-025) Does pre-slaughter transportation and lairage affect Salmonella enterica shedding prevalence
and levels in market pigs?
This research was conducted with the objective of determining the effect of pre-slaughter transportation and lairage on
Salmonella enterica shedding prevalence, shedding levels of total Enterobacteriaceae and Lactobacilli, and proportion
of the sub-populations of antimicrobial resistant Enterobacteriaceae and Lactobacilli in market pigs, under commercial
conditions. Five replicates of the experiment were conducted with 30 pigs each. Each animal was individually identified
and fecal samples collected immediately before moving the animals from their resident pens to a transport trailer (pre-
transport), immediately upon arrival at the abattoir (post-transport), and after approximately 2 hours of pre-slaughter
resting (post-lairage). Salmonella prevalence increased significantly between each one of the defined sampling points
(11.3% pre-transport, 20% post-transport, and 42% post-lairage). The total Lactobacilli population enumerated was
significantly higher than the total Enterobacteriaceae population at all sampling points. No effect on the total numbers of
total Enterobacteriaceae or total Lactobacilli was observed. No quantitative effect was observed on sub-populations of
Enterobacteriaceae resistant to ampicillin, tetracycline, and gentamicin. Also, no quantitative effect on sub-populations
of Lactobacilli resistant to tetracycline and erythromycin was observed. However, a significant quantitative increase of the
ampicillin-resistant Lactobacilli was observed from the pre-transport to the post-lairage sampling. This research suggests
that effects of stress on the microbial ecosystem of the gastrointestinal tract of swine differ between groups of bacteria,
and also between its sub-groups. Our findings reveal a critical need for further research to advance our knowledge on the
quantitative effects of stress on microbial populations of the gastrointestinal tract of swine and its potential food safety
implications.
(06-054) Quantifying Salmonella and Campylobacter Peri-harvest in Swine

Salmonella and Campylobacter are estimated to cause 3.9 million illnesses annually in the United States, and most of these
illnesses are food-related. Pigs can be sub-clinically infected with these pathogens and fecal contamination of meat during
processing is a food safety risk. Quantitative measures of foodborne safety risk are rarely reported and are a critical data
gap for development of quantitative risk assessments. The goal of this study was to determine the association between the
concentration of Salmonella and Campylobacter in porcine feces and hide with concentrations in meat. Methods: Samples
were collected 5 times from 100 individually identified pigs during the peri-harvest period. Feces were collected on the
farm and in lairage. A hide swab was collected before scalding and the entire carcass was swabbed immediately before
chilling. For each individually identified carcass a meat sample was collected. Salmonella and Campylobacter was cultured
and quantified at each stage using the Most Probable Number Method and Real Time PCR. Results: Salmonella was culture
positive from two samples (one farm sample and one lairage sample from different pigs). The proportion (%) of samples
that were Campylobacter positive for MPN method was 90, 94.6, 75.5, 100, and 48.9 for farm, lairage, hide, carcass and rib
samples respectively. The proportion (%) of samples that were Campylobacter positive for Real Time PCR was 48, 57, 68, 69,
40, 0, and 0 for fecal sample from farm, fecal sample from lairage, hide swab, carcass swab, ground meat and rib samples
respectively The mean Campylobacter concentration for each sample type for MPN method was: farm, 1,665,122.9 cfu/g;
lairage, 12,014,729.5 cfu/g; hide, 1.4 cfu/cm2; carcass, 1733.9 cfu/half carcass; and ribs, 17.5 cfu/g. The mean Campylobacter
concentration for each sample type for Real Time PCR was: fecal sample from farm, 1,665,122.9 cfu/g; fecal sample from
lairage, 12,014,729.5 cfu/g; hide swab, 1.4 cfu/cm2; carcass swab, 1733.9 cfu/half carcass; and ribs, 17.5 cfu/g. A weak
positive correlation between Campylobacter concentrations in feces (farm, R=0.2 [p=0.065] and lairage, R=0.2[p=0.068])
and concentration of Campylobacter on meat was found using the Spearman’s rank coefficient for the MPN method. No
associations were found on qualitative measures using odds ratios for the MPN method.
(03-120) Effects of egg yolk antibodies and antibiotic regimens on carriage and shedding of Salmonella
Typhimurium
Salmonella may reside in pigs during all stages of growth and may later contaminate pork products, thus presenting a
foodborne hazard for humans. Often, Salmonella are passed among hogs during transport to and holding in slaughter
facilities, thus increasing the potential for contamination of pork products. This research tested an egg yolk product that
contained anti-Salmonella antibodies as a feed additive to determine if such products can reduce Salmonella in pigs.
Experiments were conducted to determine the effectiveness of the egg yolk product in young pigs and in older hogs
bound for market. The results of these studies indicate that egg yolk antibodies are not an effective means of reducing
Salmonella in pigs, whereas diets containing some antibiotic products can reduce Salmonella in younger pigs. However,
as antibiotic use can lead to drug-resistant bacteria, the search for alternative feed additives or management schemes
that reduce foodborne bacteria should remain a priority for the industry.
(02-135) Investigating the Genetic Relatedness among Salmonella Species Isolated in Transportation
Trucks, Holding Pens and From Swine at Slaughter
Previous studies have suggested strongly that the lairage is a source of Salmonella to swine. Using more refined tools to
determine relatedness between isolates we have reinforced this finding. More importantly however it appears that some
serotypes of Salmonella may be more able to “transfer” from the pen to the pig.
(00-098) Evaluation of Antemortem Salmonella Detection Procedures in Market Weight Swine

The objective of this study was to compare fecal culture, meat juice ELISA, and culture of abattoir collected tissues
for their ability to accurately estimate the on-farm prevalence of Salmonella enterica in market swine. Six conventional
commercial herds, depopulated in the Accelerated Pseudorabies Eradication Program (APEP) were used. One-half of
study pigs (50) were necropsied on-farm; the other half (50) at a commercial abattoir. The true farm prevalence (TFP),
based on the summation of positive cultures from ileocecal lymph nodes, cecal contents, or fecal samples was estimated
at 5.3%. This estimate was higher than any provided by a single sample type cultured. The seroprevalence (meat juice
ELISA) was estimated at 20% using a cutoff of OD% > 40. Swine were transported directly to the abattoir in disinfected
trucks. They were allowed to reside in the antemortem pens for an average of 2.5 hours (1.5 to 3.5 hrs) before harvest. The
tissues sampled were the same as on-farm. Prevalence estimates from abattoir collected samples were significantly higher
than TFP; 39.9% of pigs were culture positive. Three Salmonella enterica serotypes were identified on farm. These and an
additional 14 serotypes were found in abattoir collected samples. This study shows that a single round of fecal collections
will underestimate the true Salmonella status of a herd and that samples collected at the abattoir will overestimate the
on-farm prevalence. It also demonstrates a significant rise in prevalence and serotype distribution between the production
site and abattoir. Antemortem holding pens are considered a significant risk factor for this rise in prevalence.
(98-149) Group relationship of Salmonella ELISA antibody status of grower-finisher hogs to fecal
shedding detectable by culture
From July 1998 through November 1999, 15 groups of 30 pigs (450 total) were randomly picked (each total group size =
185 pigs), double ear-tagged, individually fecal and blood sampled, and placed into one of 5 separate finishing facilities
(3 groups per facility) within a vertically integrated pork production system. Fecal and blood sampling was continued
on individually identified animals at approximate monthly intervals for 3 or 4 times with the last pre-harvest samples
being collected 2 to 18 days prior to slaughter (9 day average). All groups of hogs remained on full feed until loaded for
shipment. Transportation time and methods, separation of hog groups during transit and at the packer, and lairage were
approximately the same for all groups. Ileocecal lymph nodes and cecal/rectal combined fecal samples were collected at
slaughter from individually identified hogs. Sera was tested by the MIX-ELISA test for Salmonella antibodies and fecal and
cecal/rectal fecal samples were cultured for Salmonella, serogrouped, and sent to NVSL for serotyping.
(97-2002) Affect of Feed Withdrawal on the Shedding of Salmonella Typhimurium by Swine

This project was designed to determine if transportation-related stress and feed withdrawal caused increased shedding
of Salmonella by carrier pigs. In this set of experiments, 48 pigs were challenged orally with S. Typhimurium after weaning
and allowed to grow under typical production practices except that antibiotics were not included in feeds. At monthly
intervals, fecal and serum samples were collected from each pig. All pigs shed the challenge organism at least once during
the experiment. By the sixth month, most pigs were negative for the challenge organism when cultured from feces. When
pigs reached market weight (~240 pounds), they were split into 4 groups and subjected to the following feed withdrawal
protocol: group 1 had no feed withdrawal, groups 2-4 had feed withdrawn at 6, 12 or 24 hours, respectively. Afterward,
pigs in each group were transported ~140 miles (4 hours), returned to the production facility and necropsied. Contents at
the ileal-cecal junction were collected and cultured for the test organism. There was a direct correlation between the time
of feed withdrawal and the number of pigs with S. Typhimurium in intestinal contents. Pigs that had no feed withdrawal
had the fewest number of positive pigs, while pigs that were off feed for 24 hours had the highest number of positive
pigs. Serum samples were assessed using a mixed ELISA to detect antibodies against Salmonella. All pigs had detectable
antibodies. Over the various bleedings, most pigs had increases in anti-Salmonella antibodies over time. Some did show
modest drop-offs at the later times. However, about 10% of pigs exhibited constant titers during all times sampled. In
each case, the specific titers from these low responding pigs were always low compared to the other pigs. The results from
this experiment complement our previous work and demonstrated that both transportation stress and feed withdrawal
contribute to shedding of Salmonella by pigs. Furthermore, most pigs were not shedding the challenge organism just prior
to slaughter, yet most still had persistent infections. This indicates that the detection of infected, asymptomatic carriers will
be difficult could complicate any plans to develop programs to certify Salmonella-free pigs.
(97-1900) The Influence of Competitive Exclusion Cultures and Fructooligosaccharide Supplementation

on Salmonella in Pigs
Two challenge trials were conducted in an attempt to develop a consistent challenge model to be used for determining
effective interdiction methods to reduce shedding of Salmonella prior to slaughter. Ideally a challenge model should have
most challenged pigs shedding continuously at moderately high levels for at least one week. Neither increasing dose
level (103 to 1011), isolate used, challenge with PRRS, nor injection with dexamethazone developed a consistent shedding
state in pigs at the late finishing phase of growth. Salmonella was isolated from cecal and/or ileocecal lymph nodes of
four of six PRRS infected animals at necropsy, however Salmonella was never isolated from fecal samples from these
animals. Alternative procedures to reduce both the intestinal microbial inhibition and the animal’s resistance of pathogen
colonization need to be developed to develop a model to identify potentially effective interdiction pre-harvest methods
to reduce food borne contamination of meat products.
Section 2: Peri-Harvest
CONTROL STRATEGIES
(13-258) Further Characterization of Salmonella and Campylobacter isolates from NARMS Study
The primary objective of this research was to examine the potential impact of normal lairage practices on Salmonella,
Campylobacter and E. coli antimicrobial resistance patterns in swine following evisceration in an abattoir environment.
Fecal samples were collected upon arrival at the abattoir and later from the cecum within cohort groups. Also non-
cohort pig and environmental samples were collected from the abattoir for comparison purposes. Isolates of Salmonella,
Campylobacter, and E. coli tested for resistance to a panel of antimicrobials as well as characterized by serotype,
phenotype and genetic patterns as appropriate. The results indicate a wide variety of antimicrobial resistance and other
patterns that were not consistent between arrival and post slaughter samples. The conclusion is that normal lairage
practices do impact bacterial populations in the gut in a manner which alters their antimicrobial resistance profiles. Thus
resulting antimicrobial resistance patterns from samples collected post slaughter are not useful measures for monitoring
on-farm resistance since they are even farther removed from the source then arrival samples. These results will aid the
swine industry and inform research and regulatory communities as they seek to develop and enhance surveillance systems
which reflect the impact of on-farm antimicrobial use.
(09-131) Phage therapy to reduce transport and lairage associated increases in Salmonella infections:
evaluation of delivery methods
Bacteriophages are naturally occurring microorganisms that target and destroy bacteria. Our laboratory previously
demonstrated that administering bacteriophages to pigs reduces Salmonella colonization when those pigs are exposed
to a contaminated environment. These results indicated that bacteriophages could be useful in controlling increases in
Salmonella shedding that occur just prior to processing due to transportation and lairage. The purpose of the current
project was to increase the efficacy of our phage treatment by: 1) increasing the host-range to include other common
Salmonella serovars associated with swine; 2) developing a more affordable microencapsulation process to protect
the phages from stomach acids; and 3) making deliver more practical by determining whether phages can be direct
fed. We collected samples from wastewater treatment facilities and isolated 20 distinct phages belonging to either the
Siphoviridae or Myoviridae families. From this library we identified 10 phages that together lysed a mixed culture of
Salmonella enterica Typhimurium, Enteriditis, and Kentucky. Phages were microencapsulated using two methods that had
minimal effect on phage viability. Microencapsulated phages remained stable at various temperatures for up to 14 days
with no appreciable drop in viability. Twenty-one pigs in three replicates were randomly placed into three groups: feed (F),
gavage (G), and control (G). The F group was direct fed the phage cocktail daily for five days. On the fifth day, the G group
received the same phage cocktail by gavage, while C pigs received a mock treatment with no phage. All pigs were then
challenged with Salmonella enterica Typhimurium. Pigs in the F group were significantly less likely to shed Salmonella
post-challenge compared to pigs in both the G and C groups. Likewise, the concentrations of Salmonella in the ileal and
cecal samples were significantly lower than ileal and cecal samples from C pigs. Taken together, these results indicate that
direct feeding phages is a practical and effective means of reducing Salmonella colonization and shedding in pigs.
(06-167) Employing phage therapy to reduce lairage associated increases in Salmonella infections and
shedding
Contamination of meat and meat products with foodborne pathogens is usually the result of the carcass coming
in contact with the feces of an infected animal during processing. In the case of Salmonella, several recent studies
have reported that pigs become rapidly infected with the organism during transport and lairage due to Salmonella
contaminated trailers and holding pens. These infections serve to increase the likelihood of carcass contamination by
amplifying the amount of bacteria that enters the processing facility. To counteract these infections, we have developed
a phage based anti-Salmonella intervention strategy that can be administered to pigs just prior to transport and lairage.
We have isolated and characterized 15 wild-type, Salmonella-specific phage and developed a means be which the viruses
can be microencapsulated for administration to animals. Initial tests of the efficacy of our anti-Salmonella phage cocktail
were conducted in small pigs (~20-30 pounds) where we treated the pigs with the phage cocktail and then inoculated
them with Salmonella enterica Typhimurium. Infections in pigs administered the phage treatment were reduced 99.0%
to 99.9% in the tonsils, ileum, and cecum. To test our anti-Salmonella phage cocktail in a more production-like setting,
we infected (in three replicates) four market weight pigs with Salmonella enterica Typhimurium. These seeder pigs were
allowed to contaminate a holding pen for 48 hours. At 48 hours post-inoculation, 16 naïve pigs (in three replicates) were
introduced to the contaminated environment. Eight pigs were given the anti-Salmonella phage cocktail while the other
eight were mock-treated. Phage treatment reduced cecal and ileal Salmonella infections by approximately 95% and 90%,
respectively, in phage treated pigs. These data indicate that phage therapy can be used as an effective means to reduce
or limit Salmonella infections during the crucial periods of transport and lairage.
(06-086) Quantitation of Salmonella in transport and lairage to assess interventions for reduction of
cross-contamination in pigs
The sensitivity of the quantitative polymerase chain reaction assay (qPCR) was shown to be 10,000 Salmonella colony forming
units of bacteria per gram of fecal matter; this means that this rapid and specific test is a useful tool for the industry. The
study found a trend toward reduced, but not statistically significant, amount of Salmonella in lairage pens from standard
hygiene practices of washing pens between groups of swine. This reduction in lairage pens may reduce transmission
between groups and could result in a reduction of contaminated carcasses. It also allows one to determine if on farm and
transport sanitation methods are reducing Salmonella fecal incidence at lairage pens. The lack of significance is partially due
to the low number of positive pens in this study before washing. Due to the low numbers found in this study, samples were
not collected from transport as they would have not lead to significant differences between various treatments.
(03-136) Pre-nitrate adaptation and chlorate supplementation to reduce Salmonella, Escherichia coli,
and Yersinia enterocolitica in swine
Salmonella, E. coli and Yersinia are bacteria that can cause foodborne illness in humans. These bacteria can be found in
the intestinal tract of pigs, therefore methods are sought to rid these bacteria from pigs before slaughter. We previously
demonstrated that an experimental chlorate product could reduce Salmonella and E. coli in pigs. In the present study,
certain nitrocompounds were able to kill Salmonella, E. coli and Yersinia in the laboratory and when fed to swine, markedly
enhanced the bacterial-killing capacity of the chlorate product. Subject to regulatory approval, it is possible that these
compounds may be developed into feed additives that can be fed prior to slaughter to rid bacterial pathogens from swine.
(02-129) Eliminating the Abattoir Pen Lairages to Decrease the Prevalence of Salmonella in Cull Sows
This project has demonstrated that the antemortem holding pens at an abattoir may play a role in increasing the
number of Salmonella-infected cull sows entering the abattoir. It also provides an indication that other peri-marketing
concentration and transport steps may be involved in this contamination, as exhibited by the heightened isolation rates
of the direct harvested (no hold) groups when compared to levels expected in farm-direct marketed swine. Exposure to
contaminated facilities for as short as a 2 hour period may raise these isolation rates. The cecal contents demonstrated
the only significant rise in all the tissues cultured. This is consistent with other field and laboratory exposure observations.
The failure to culture Salmonella from carcass swabs or the meat blocks produced from the test animals demonstrates that
it is possible to reduce Salmonella contamination of meat products by implementing appropriate butchering techniques
and other HACCP-based control steps at the packing plant level.
(02-119) Vaccination to Prevent Acute Infection by Salmonella in Transport and Lairage Prior to Slaughter
Currently available Salmonella Choleraesuis live avirulent vaccines are efficacious for the prevention of disease caused by
S. Choleraesuis. This research was conducted to determine if these same vaccines could prevent infection by Salmonella
Typhimurium, an important human foodborne pathogen present in pigs pre-harvest. These results indicate that S.
Choleraesuis avirulent vaccines do not prevent the rapid dissemination of S. Typhimurium when pigs are challenged
either with a high dose (intranasal) or a low dose (natural challenge via pen contamination) of the organism. It is very likely
these avirulent vaccines do decrease the levels of Salmonella in pigs at slaughter but this protection occurs when pigs are
exposed to Salmonella during the finishing phase of production, not during transport and lairage. Treatment of pigs with
spray-dried plasma does not prevent the rapid dissemination of S. Typhimurium.
(02-108) Evaluation of the Efficacy of Various Intervention Methods Used by Small Processors for Pork
Carcasses Contaminated with Salmonella spp.
The main objective of this experiment was to determine the most efficient S. Typhimurium intervention method suitable
for small scale industrial implementation. Ultimately, the specific intervention method used did not matter as no
distinguishable differences were observed between inoculated carcasses that received different intervention treatments,
a standard hot water rinse and 24 h chill step. A general intervention should be implemented following dehairing and
evisceration because both processes present the possibility for re-contamination. The hot water (53ºC) sprayed at 70
psi was the most effective individual intervention but it was essential that the carcasses also be rinsed or receive some
antimicrobial treatment after evisceration and chilled to achieve a near complete reduction of 8.0 Log CFU/cm2 S.
Typhimurium. Also, some caution should be heeded when applying high pressure sprays during swine slaughter as the
jowl area was found to be the least accessible area to these treatments. The combination of various decontamination
strategies during livestock slaughter is encouraged by the FSIS as they have been proven to reduce enteropathogen
populations on carcasses (Fed. Register, 1995).
Salmonella spp. prevalence since HACCP was implemented has dropped mainly due to several meat and poultry
establishments incorporating antimicrobial treatments during processing in order to meet FSIS performance standards
(Rose et al., 2002). There is an abundance of research currently being conducted with the objective of finding the most
efficient combination of antimicrobial treatments while in most cases, neglecting the significance of the chill step. The chill
step had the greatest influence on the S. Typhimurium strains used in this experiment.
(02-074) Effect of Chlorate Treatment on Transmission of Salmonella in Swine during Transport and Lairage
Chlorate can be used to reduce Salmonella in pigs. However, it must reach the lower gut in order to kill Salmonella.
Therefore, although chlorate will work to reduce Salmonella during the lairage period, it must be fed to the animal
approximately 16-24 h prior to slaughter in order to reap any benefit from chlorate treatment. Thus, we cannot
recommend the use of chlorate as a treatment specifically in the short-term (6 h) lairage period, although it does appear
beneficial when used a short period of time (16-24 h) before slaughter. Additionally, these results indicate that proper
facility sanitation is crucial to the effectiveness of this, and any, intervention strategy. We thank the National Pork Board
and our nation’s pork producers for their generous financial assistance in exploring this important avenue of improving
food safety.
(00-136) Effect of Sodium Chlorate on Salmonella Typhimurium in the Pig Gut

Salmonella cause disease and compromise food safety. Consequently, strategies are sought to reduce their concentration
in pigs immediately before processing. Respiratory nitrate reductase activity possessed by Salmonella coincidentally
catalyzes the intracellular reduction of chlorate to chlorite, a consequence that kills the microbe. Since most beneficial gut
bacteria lack respiratory nitrate reductase, we conducted several studies to see if chlorate may selectively kill Salmonella,
but not beneficial microbes, within the pig gut. In the first study, weaned pigs orally infected with Salmonella Typhimurium
were treated 8 and 16 h later via oral gavage (10 ml) with 0 or 100 mM sodium chlorate. The pigs were euthanized at 8 h
intervals after receiving the last treatment and samples collected by necropsy were cultured for Salmonella. A significant
chlorate treatment effect (P < 0.05) on cecal Salmonella concentrations was observed, although a treatment x time
after treatment interaction was also observed which suggests that the chlorate effect was concentration dependent.
In follow up studies, 0, 15 or 30 mM sodium chlorate solutions were administered via drinking water to weaned or
finished pigs experimentally challenged 24 h earlier with Salmonella Typhimurium. These pigs were euthanized 12 h or
24 h after being allowed ad libitum access to their respective treatments and gut samples were again cultured for the
challenge Salmonella strain. In these studies, Salmonella concentrations in gut contents collected from chlorate treated
pigs were reduced up to 1000-fold compared to concentrations from control pigs. These results demonstrate that
chlorate administration may be a practical procedure for reducing gastrointestinal concentrations of Salmonella in pigs
immediately prior to processing.
Section 3: Post-Harvest
PREVALENCE IN PORK AND
PORK PRODUCTS
(14-203) Prevalence and characterization of Salmonella from Head Meat and Trim for Ground at Pork
Processing Facilities
Recent research has reported that if superficial lymph nodes from cattle carcasses were included into ground beef,
there was an increased risk of Salmonella contamination of ground beef. This increased risk of contamination is because
Salmonella is often at a higher prevalence in lymph nodes compared to other tissues. Pork products, particularly the head
trim and other trim destined for ground pork, may contain lymph nodes. It is not unreasonable to speculate that pork, like
beef, could be at increased risk of Salmonella contamination if lymph nodes are present in ground pork. To date, there
has been limited to no research on Salmonella prevalence for chops and roasts, head trim, or trim intended for ground.
We had preliminary data from a pork processing plant to indicate that a high percentage (98%) of cheek meat can be
contaminated with Salmonella. The objectives of the present study were to: 1) determine the prevalence of Salmonella
in head meat and trim intended for ground; 2) determine the serotypes (genetics) of isolates; 3) determine the antibiotic
resistance of isolates; and 4) if justified, use molecular techniques to determine the relatedness of isolates that are of the
same serotype, but display differences in antimicrobial patterns.
In this study, a large pork processing plant in the United States was sampled every other month for 11 months from
January to November of 2015 to determine the prevalence, seasonality, diversity of serotypes, and antimicrobial
susceptibility of Salmonella enterica isolated from cheek meat and head trim of swine carcasses. Each cheek meat and
head trim collection period (January, March, May, July, September, and November) consisted of 25 samples collected
on a Monday a.m., 25 on Monday p.m., 25 on Tuesday a.m., and 25 on Tuesday p.m., for a total of 100 cheek meat and
100 head trim samples (total of 200 for each period, total of 1200 for 6 periods). Tissues were cultured for Salmonella by
described procedures using restrictive media and enrichment techniques. Salmonella isolates were serotyped by the
National Veterinary Services Laboratories, Ames, IA, USA.
For the six sample periods, the percentages of SE-positive sample totals were 63% for cheek meat and 66% for head trim
for a total of 774 isolates. The following were the results of isolations from cheek meat and head trim: January 94%; March
80%; May 53.5%; July 58.5%; September 46.5%; and November 55%. There was a large diversity of serotypes (25) which
included isolates commonly found in swine and others that have rarely seen in swine. We identified 218 isolates (99 [58.8%]
cheek meat and 119 [66.8%] head trim) that were multi-drug resistant (greater than 3 classes of antibiotics). Of the multi-
drug resistant isolates, 90 were the type identified by a national program (National Antimicrobial Monitoring System) as
Salmonella (ACSSuT phenotype) of increasing concern. Additionally, increased resistance to ciprofloxacin, a broad spectrum
antibiotic used in human medicine, was observed in isolates and was attributed to genetic elements within the bacteria.
The results from this study suggest that pork products from the head have a high carriage rate of Salmonella which
includes a diverse population of serotypes with a substantial number of isolates with elevated multi-drug resistance.
Based on our results, there appears to be an effect of season with increased prevalence of Salmonella in cooler months
(Jan., Mar., Nov. = 76.3%) compared to warmer ones (May, Jul., Sept. = 52.8%). The results from this study are beneficial
to the industry because now there is a knowledge base of the extent of the Salmonella prevalence, the level of antibiotic
resistance in Salmonella from swine products, the potential seasonality of Salmonella carriage in swine, and the wide
range of serotypes and genetic diversity of Salmonella in swine products. With an increased knowledge of a problem
comes the search for solutions. Intervention methods to reduce Salmonella in head products of swine processing plants
are warranted and will be forthcoming. Overall pork wholesomeness will be improved and thus a “value added product”.
(12-170) Determining the incidence and antimicrobial susceptibility of Salmonella serovars isolated
from U.S. retail ground pork
Salmonella is a foodborne pathogen that may be associated with meat products. The goal of this project was to
determine the prevalence of Salmonella in ground pork at retail, and to evaluate factors that may be associated with
its prevalence. Because ground pork at retail can be regulated by two separate governing bodies, the authors wanted
to measure Salmonella prevalence in case-ready packages versus store-ground pork using USDA and FDA Salmonella
isolation protocols. Finally, the Salmonella isolated from the ground pork was serotyped and evaluated for antimicrobial
resistance using 15 antibiotics.
Ground pork was collected from grocery stores, supermarkets, and retail establishments that sold fresh meat. Packages
were collected from 12 cities in 4 different regions of the U.S. (South, East, West, and Central/Midwest) during 3 seasons
(Fall, Winter/Spring, and Summer). Approximately one-half of the packages collected in each city were ground and
packaged off-site (referred to as “case-ready”) and the other half were ground on-site (referred to as “store-ground”).
Packages were tested for Salmonella using USDA-FSIS and FDA Salmonella isolation protocols, as well as genomic
evaluation. Salmonella isolates from ground pork were additionally evaluated for antimicrobial resistance using 15
antibiotics. Finally, all Salmonella isolates were serotyped.
Overall, 1.39% of ground-pork packages were positive for Salmonella (12 of 865 packages). There was no difference in
Salmonella prevalence between case-ready and store-ground pork or between package types (overwrap, chub, MAP, or
other). More Salmonella was isolated during the Fall season that any other season, and there was a tendency for increased
prevalence in the East region. The USDA-FSIS method was a more effective method to isolate Salmonella compared to
the FDA isolation protocol. None of the isolates were resistant to antibiotics used to treat Salmonella infections, such as
extended spectrum cephalosporins or flouroquinolones. Six different serotypes were isolated, and only two packages
contained multiple Salmonella serotypes.
The results of this study indicate the observed prevalence of Salmonella in ground pork products was low. Moreover, a
clear majority of the Salmonella serotypes were broadly susceptible to the antibiotics tested and none of the serotypes
were resistant to critically important antibiotics used in human medicine. Finally, these results indicate the pork industry
should continue to utilize technologies and practices to reduce pathogens in pork products.
(12-145) National Prevalence of Salmonella Contamination in Retail Ground Pork

Salmonella is a foodborne pathogen that may be associated with meat products. The goal of this project was to
determine the prevalence of Salmonella in ground pork at retail, and to evaluate factors that may be associated with
its prevalence. Because ground pork at retail can be regulated by two separate governing bodies, the authors wanted
to measure Salmonella prevalence in case-ready packages versus store-ground pork using USDA and FDA Salmonella
isolation protocols. Finally, the Salmonella isolated from the ground pork was serotyped and evaluated for antimicrobial
resistance using 15 antibiotics.
Ground pork was collected from grocery stores, supermarkets, and retail establishments that sold fresh meat. Packages
were collected from 12 cities in 4 different regions of the U.S. (South, East, West, and Central/Midwest) during 3 seasons
(Fall, Winter/Spring, and Summer). Approximately one-half of the packages collected in each city were ground and
packaged off-site (referred to as “case-ready”) and the other half were ground on-site (referred to as “store-ground”).
Packages were tested for Salmonella using USDA-FSIS and FDA Salmonella isolation protocols, as well as genomic
evaluation. Salmonella isolates from ground pork were additionally evaluated for antimicrobial resistance using 15
antibiotics. Finally, all Salmonella isolates were serotyped.
Overall, 1.39% of ground-pork packages were positive for Salmonella (12 of 865 packages). There was no difference in
Salmonella prevalence between case-ready and store-ground pork or between package types (overwrap, chub, MAP, or
other). More Salmonella was isolated during the Fall season that any other season, and there was a tendency for increased
prevalence in the East region. The USDA-FSIS method was a more effective method to isolate Salmonella compared to
the FDA isolation protocol. None of the isolates were resistant to antibiotics used to treat Salmonella infections, such as
extended spectrum cephalosporins or flouroquinolones. Six different serotypes were isolated, and only two packages
contained multiple Salmonella serotypes.
The results of this study indicate the observed prevalence of Salmonella in ground pork products was low. Moreover, a
clear majority of the Salmonella serotypes were broadly susceptible to the antibiotics tested and none of the serotypes
were resistant to critically important antibiotics used in human medicine. Finally, these results indicate the pork industry
should continue to utilize technologies and practices to reduce pathogens in pork products.
(10-146) Evaluating the sources of Salmonella after carcass chilling

This study evaluated several aspects of Salmonella contamination of pork after slaughter but prior to retail. First, a
comparison was made between the USDA-FSIS carcass swab procedure and two other sampling methods. The FSIS carcass
swab had the lowest level of sensitivity in comparison to excision samples and the M-Vac sampling device. This suggests
that the carcass sponge swab method may be providing less accurate results, leading processor to miss early signs of
processing deviations. A second comparison was made between the prevalence of Salmonella as determined by carcass
swabs, loin swabs and trim samples in a commercial establishment. The data shows that Salmonella on finished carcasses
was rare, but when it did occur, it appeared to occur frequently within carcasses from the same herd. A higher incidence
on carcasses was also reflected in positive loin samples. Finally, a series of experiments were conducted to evaluate the
potential for conveyor belts to be a potential source of cross contamination. These studies showed that, irrespective of belt
material type, conveyor belts could be a potential source of cross contamination, even at low populations.
(08-173) White Paper on Human Illness Caused by Salmonella from all Food and Non-food Vectors
Salmonellosis continues to be an important foodborne disease in the U.S. and worldwide. Even though illness caused by
Salmonellae is usually not very severe, there are such a large number of cases that salmonellosis has an estimated five fold
greater economic impact than illness caused by E. coli O157:H7.
Overall objective of this project was to summarize historical and epidemiological data published in the scientific literature
on the relationships between food and non-food vehicles and vectors and human illnesses caused by Salmonella.
Scientific literature databases, U.S. government publications, relevant government publications/regulations from
other countries, industry information and other internet sources were searched for information on outbreaks and
epidemiological studies of human illness attributed to Salmonella spp. until December 2008. Data were also collected on
important government regulations and industry initiatives to control this organism.
Information was presented in a 51-page white paper with 485 references describing epidemiology, surveillance,
regulation, and industry initiatives to control Salmonella. Data on 501 outbreaks (1992-2008) were tabulated
chronologically in an appendix to the white paper. Outbreak data were also organized in tables presenting information on
the largest outbreaks (>200 cases), vehicles of infection associated with different Salmonella serotypes, and Salmonella
serotypes associated with different vehicles. Figures were generated to display importance of different vehicles.
Data were discussed in the text with reference to: (a) reservoirs of these bacteria, (b) routes of human infection, (c)
relationships between food and non-food vehicles and vectors, (d) factors important in causing illness, (d) government
regulations, (e) industry initiatives and (f) surveillance strategies.
Of the outbreaks with known or suspected vehicles, 37% of outbreaks and 34% of cases were traced to meat and meat
products. Some other outbreaks associated with combination foods (lasagna, sandwiches, pot pies, etc.) may also have
been caused by contaminated meat. Of the meat outbreaks, pork was associated with 30% of outbreaks, 36% of cases. In
addition to outbreaks associated with pork meat, ham, and sausage, there were outbreaks among people handling pig
ear dog treats. S. Typhimurium was by far the most common of the 12 serotypes associated with pork products. Poultry
products were associated with 33% of meat outbreaks and beef products with 18%. Outbreaks and cases of salmonellosis
associated with fresh fruits and vegetables have become more common.
About 65% of Salmonella isolates from turkeys and swine are resistant to at least one antibiotic compared to 30-40% of
isolates from chickens and cattle. Data indicate that prevalence of antibiotic sensitivity in Salmonella from turkeys, swine,
and cattle has decreased between 1999 and 2006. However, resistance to multiple antibiotics has increased, particularly in
S. Typhimurium and S. Newport.
Some industry interventions, such as improved hide cleaning processes, have reduced Salmonella contamination of
animal carcasses. Some European countries have undertaken well organized, nationwide programs to control Salmonella
in poultry and swine.
(07-050) Enumeration of Salmonella throughout the Pork Harvesting Process

This project provides comprehensive information on the levels and prevalence of Salmonella on pork carcasses during
processing to benchmark the effectiveness of in-plant interventions on Salmonella reduction. During this project, 190 pork
carcasses were sampled over two consecutive days in each of four seasons at a pork processing plant in the United States.
Samples were collected at three steps in processing: skin (after exsanguination and before scalding), pre-evisceration
carcass (after scalding, singeing, and polishing) and chilled final carcass in the cooler. Overall prevalence of Salmonella
on skins was near 100% in all seasons with a one-day low of 85% in the winter. The range of Salmonella prevalence
on pre-evisceration carcasses was between 10.5% and 69.5%. Final carcasses had 12.6% Salmonella prevalence in the
summer, undetectable (0% prevalence) in the fall, ~4.2% prevalence in the winter and ~9% in the spring. The prevalence
of Salmonella was twice as high on skins and three times higher on pre-evisceration and final carcasses as was expected.
During the four seasons, the lowest enumerable level of Salmonella bacteria on skins was 27 cfu/100cm2 and the highest
level was 2322 cfu/100cm2. Salmonella serotype and drug resistance patterns have been determined for 2,176 isolates.
From this data set, 22 serotypes have been identified and the antibiotic profiles of these range from having resistance to
eight antibiotics to being susceptible to all 14 antibiotics tested. When the amount of Salmonella on skins was high, as
indicated by both a large number of enumeration positive samples and high enumeration levels, the chance of having
Salmonella on carcasses was higher. However, the percentage of carcasses with levels high enough to enumerate and the
enumeration levels were both relatively low.
(06-186) Baseline Study - Prevalence and Levels of Salmonella and Campylobacter on Retail Pork Cuts
A total of 4,000 retail pork products were examined during January – March, 2007, from 4 locations in the Midwestern and
Southwestern United States, including Green Bay, WI, Dallas, TX, San Antonio, TX and Phoenix, AZ. Four representative
pork products were collected during this study, including enhanced pork chops, enhanced boneless pork roasts, un-
enhanced (natural) pork chops and un-enhanced (natural) boneless pork roasts. Pork products were collected in a similar
manner at each location for the first two-thirds of the sampling period. For the final, one-third of the collection period only
enhanced products primarily from the Phoenix, AZ location were sampled based on the inability to recover Salmonella
spp. from un-enhanced products, and further, on the limited number of positive samples recovered in enhanced products
at the other 3 locations. Retail samples were first screened for the presence of Salmonella spp. in sets of 5 per product
type. If the sample set screened positive, further testing was performed on the individual samples associated with the
set. In the event that an individual sample was determined to be positive for Salmonella spp., it was assessed for the
estimated level of Salmonella spp. present per gram of product.
CONTROLLING MULTI-DRUG RESISTANT
(MDR) BACTERIA IN PORK
(16-123) Assessing the unintended effects that therapeutic antibiotic treatment has on multidrug-
resistant Salmonella in swine: implications for swine health and food safety
(07-200) Impact of organic acids and sanitizers on Salmonella serovars with and without SGI1-mediated
multi-antibiotics resistance
Multidrug-resistant (MDR) Salmonella has been implicated in several infections and outbreaks worldwide; they have
been linked to the consumption of contaminated products involving pork as well. Despite the interventions of HACCP
in abattoirs, Salmonella is still isolated from carcasses, and retailers (NARMS, 2005). The objective of this study was to
compare the relative efficacy of organic acids and sanitizers against multiple drug-resistant (MDR) serovars of Salmonella
and non-resistant Salmonella in order to gain more information on the potential effectiveness of intervening chemical
treatments in pork slaughter and processing. The study also determined if exposure of normal and MDR Salmonella to
organic acids affects sensitivity to sanitizers. Observed changes in Salmonella isolates which either had or did not have the
qacE∆1 gene were compared.
This research was conducted in four steps. First, MDR-Salmonella were exposed to two different organic acids (lactic and
acetic acid) with samples withdrawn at different times to determine survival due to acid exposure. Bacteria were first pre-
adapted to acid conditions, since it has been found that acid adapted bacteria are more resistant to further treatments.
The second stage of the project was performed by exposing the MDR-Salmonella to three different quaternary
ammonium compounds (QACs), two pure compounds (benzalkonium chloride, cetylpyridinium chloride) and one
commercial compound (S.S.4) to determine survival due to exposure to these sanitizers. Samples were taken at different
exposure times and neutralized in DE neutralizing buffer before enumerating by plating. The third phase of the project
involved the possibility of organic acids conferring cross-protection against further treatment with quaternary ammonium
compounds. This objective was developed using the same strains and QACs from the previous objectives. Strains were
pre-adjusted to acid conditions in the same way that it was done in the first objective of the project and subsequently
treated with QACs. Samples were withdrawn after 10 minutes of treatment and neutralized in DE buffer and plated in
order to determine the survival rates. The last part of the research involved the susceptibility of MDR and non-MDR
Salmonella in biofilms to treatment with quaternary ammonium compounds. Briefly, planktonic and biofilm cells were
prepared according to the procedure of Ren and Frank (1993) with slight modifications and subsequently treated with
QACs. Survival rates were determined.
Acid adapted and non-adapted Salmonella were both sensitive to the organic acid treatments; however, the acid adapted
Salmonella were more resistant than the non-adapted when challenged with lactic and acetic acid 2%. It is known, acid
adaptation can induce cross-protection to subsequent treatments such as organic acids, heat, osmotic pressure and
oxidizers (Davidson and Taylor, 2007; Greenacre et al., 2006; Foster 1991). However, acid adaptation did not appear to
induce cross-protection to further quaternary ammonium compounds treatment.
Quaternary ammonium compounds were more effective after 600 sec at 200 ppm. There was no significant difference
in response to QACs between MDR and non-MDR Salmonella. MDR and non-MDR Salmonella in biofilms were more
resistant to QACs treatment than planktonic cells, but the response to QAC treatments did not vary for MDR or non-MDR
cells when in biofilms.
According to our findings we concluded that treatments that are available in food processing plants are equally effective
against MDR and non-MDR Salmonella.
DIAGNOSTIC TOOLS
(08-120) Comparative Evaluation of Rapid Methods for Salmonella Detection in Pork

The goal of this study was to develop a single-step assay which combined all of the steps of conventional culture of
Salmonella into a single well of a 48-deep well microtiter plate. The second goal was to screen hog carcass samples
and ground pork to continuously refine the assay as it was being developed. Finally, the specificity and sensitivity of the
assay was compared to conventional microbiological Salmonella testing methods. Conventional culture begins with pre-
enrichment, followed by transfer of an aliquot of this nonselective enrichment to a selective broth containing novobiocin,
and then plating to a selective agar, such as XLT-4, on which Salmonella appear as black colonies. The components used
for conventional Salmonella isolation were consolidated into each well (5 ml/well capacity): XLT-4 was first added to each
plate and allowed to solidify after which Modified Semisolid Rappaport-Vassiliadis (MSRV) with novobiocin was added
followed by the selective broth. The 10% suspension of the sample to be screened was the final upper layer. Plates were
sealed, incubated (42°C, 48 hrs), and scored as positive if the lower XLT-4 indicator agar layer turned black. The assay
is called the RX plate because it incorporates Rappaport-Vassiliadis modified semisolid agar (R) and for XLT-4 agar (X).
Pure cultures were initially screened to ensure that the assay could detect Salmonella. Next, hog carcass lymph nodes
harvested after the deep chill were screened (n=264). This was done to ensure that non-specific blackening of the wells
due to the tissue itself did not occur. Swine lymph nodes harvested prior to the chiller, hog fecal samples (n=80) and retail
pork sausage (n=240) were also evaluated. The reliability of the RX plate was based on specificity and sensitivity estimates,
which compared RX plate results with those obtained by conventional bacteriological isolation. The RX format, while
theoretically attractive, failed to achieve the either the specificity or sensitivity of conventional microbiological culture
(08-007) Novel molecular approaches for the rapid detection of Salmonella from pork products and the
pork processing environment
Novel rapid, sensitive, and simple detection assays that can be routinely used to prevent the transmission of Salmonella in
the pork industry are in high-demand. Reverse-transcriptase Loop-mediated isothermal Amplification (RT-LAMP) assay is
a novel molecular method that has advantages over real-time reverse-transcriptase Polymerase Chain Reaction (RT-PCR)
in that it does not require expensive equipment such as a real-time PCR machine as the reaction occurs in a waterbath at
one temperature and detection is by turbidity or fluorescence after 2 h. The objectives of this research were to develop
and apply novel assays such as RT-LAMP and RT-PCR to pork products and the pork environment for the rapid and
sensitive detection of Salmonella Typhimurium and for comparison to traditional cultural methods. Another objective
was to apply the RT-LAMP assay to detect Salmonella from pork products obtained from grocery stores and from
pork processing facilities to test the suitability of the novel assays for routine testing by the pork industry in real-world
scenarios. Twenty-five gram pork chop, pork sausage, and ground pork samples were spiked with S. Typhimurium at high
(108 to 106 CFU) and low (105 to 100 CFU) inocula levels. Carcass rinse samples (500 ml; that were concentrated through
sequential filtration) and carcass swabs (100 cm2) were spiked with only high inocula levels. Samples were stomached
in 225 ml of Tetrathionate (TT) broth, portions were serially diluted and plated on Xylose Lysine Tergitol (XLT4) agar for
traditional cultural detection and RNA was extracted and assayed by RT-PCR and RT-LAMP using previously described
primers. Each experiment in duplicate was replicated at least twice.
The Trizol® RNA extraction method provided improved RNA quality over the Qiagen RNA extraction method and was
used in these studies. The RT-PCR assay on pork products spiked with high inocula S. Typhimurium showed detection
of 106 CFU/25g within one 8 h shift. Selective enrichment in TT broth for 10 h was necessary to obtain detection of 101
CFU/25g for pork chop and pork sausage, which required at least two 8 h work shifts. The previously described LAMP
assay was developed into a RT-LAMP assay that gave detection sensitivities of 101 CFU/ml after gel electrophoresis for
S. Typhimurium in pure culture and was better than RT-PCR by 1 log CFU. Spiked pork chop and pork sausage without
enrichment gave detection sensitivities of 106 CFU/25g similar to traditional plating and RT-PCR. The lower inocula levels
required selective enrichment in TT broth, to obtain detection limits of 102 CFU/25g by RT-LAMP which was one-log less
sensitive than RT-PCR and traditional plating. However, this RT-LAMP method is faster than RT-PCR by ~2 h. For both
assays, negative controls including sterile water, sterile TT broth, and un-inoculated samples did not show any positive
results, eliminating cross-reactivity or false positives. Background flora and autoclaved Salmonella cultures inoculated on
to pork samples did not show any positive results as well, again eliminating false positives, indicating the robustness of
these assays. Pork products spiked with cold stressed cells to simulate conditions associated with storage and transport
gave detection limits of ~102 CFU/25g, after pre-enrichment for 3 h in buffered peptone water and selective enrichment
in TT broth for 10-12 h. Screening of 57 natural samples from pork processing facilities and grocery stores resulted in 9
positives by traditional cultural plating, 5 positives by RT-PCR and 10 positives RT-LAMP assays. These results indicate that
RT-LAMP shows potential for application in routine testing of Salmonella from pork products that can obtain results within
two 8 h work shifts, being faster and simpler than RT-PCR and traditional cultural assays, but less sensitive by 1-log CFU.
Further work using fluorescence dyes is necessary to improve detection sensitivity that will also allow conversion of the
RT-LAMP to a quantitative assay by using automated portable fluorometers. This will further simplify the assay for routine
use by the pork industry.
A cost-effective, rapid, sensitive, Salmonella detection assay was developed for routine testing of pork commodities
that can obtain results within two 8 h work shifts. Using this sensitive portable assay that will allow rapid detection will
help to prevent contaminated products from being released in the market and contaminated products can also be
isolated to prevent cross contamination of other pork commodities and food contact surfaces. The pork industry will be
economically and socially benefited as early detection will prevent expensive recalls and associated litigation costs and
will protect brand name. Also, any processing areas or equipment found contaminated with Salmonella can be attended
to immediately, properly cleaned and improved mitigation strategies and HACCP plans can be implemented. Use of this
novel rapid assay in routine testing and surveillance will therefore aid in the prevention of Salmonella transmission by the
pork industry as well as pork-related Salmonella outbreaks in order to protect public health.
(02-027) The Development of a Novel Immunosensor to Detect Salmonella

A novel biosensor to detect Salmonella is being developed. The optical sensor is based on fluorescence resonance
energy transfer (FRET). Antibodies are labeled with fluorescent dyes that will change their fluorescence based upon
binding of Salmonella antigens to the Salmonella antibodies. A detection limit of 2.0 µg/ml was achieved for in-solution
measurements. The labeled antibodies were also immobilized onto optical fibers and then exposed to specific and
nonspecific antigen. The benchtop sensor displayed the limit of detection of 103 cells/ml.
The new immunosensor developed by this research will be utilized as an efficient and accurate alternative to monitor
bacterial contamination in pork. Such a system will be able to reduce the large economical burden caused by product
recalls and medical treatments. This biosensor is portable and will be easy to use, especially for pork producers who
do not have knowledge of food microbiology. It will permit a rapid on-site monitoring (during slaughter) system for
pork product safety. In addition to the benefits that could be realized by the pork industry, a broader impact is the
development of new diagnostic sensor arrays that could improve in the quality of food monitoring systems.
RISK ASSESSMENTS
(10-130) Risk informed management of Salmonella in deep tissue lymph nodes

The objectives of this project were to build and parameterize a quantitative risk model of the Salmonella prevalence along
the pork production chain, and therefore to apply the model in evaluating the relative impact of potential Salmonella
source, especially lymph nodes, to human food-borne risk.
Attention has turned to the potential negative role of deep tissue lymph nodes in Salmonella contamination in pork
products, such as ground pork. Previous research has shown a low likelihood of recovery from non-visceral lymph nodes
(deep tissue). The occasional recovery of Salmonella from these tissues does not necessarily represent a significant human
illness risk. However, needed is a method that could assist decision makers in quantitatively evaluating the relative impact
deep tissue lymph nodes on Salmonella contamination in pork products, thereafter the impact on public health.Without
a firm understanding on the impact, regulators may enforce further control measures to remove deep tissue lymph nodes
during swine carcass harvest, which might actually a waste of time and money.
Quantitative risk assessment is an approach that is able to quantify and compare the impact of potential risk factors on
outcomes of interest. We developed a risk assessment model quantitatively describing the Salmonella distribution and
dissemination from chilled swine carcasses to fresh ground pork serving. Deep tissue lymph nodes were simulated in the
model as a potential source of Salmonella contamination in ground pork. The other targeted source is contaminated
carcass surface. A scenario analysis was used to estimate the reduction in Salmonella contamination of ground pork if
deep tissue lymph nodes were removed and if carcass surface was decontaminated. In the scenario analysis, the amount
of deep tissue lymph nodes and the proportion of contaminated carcass surface included in ground pork were tentatively
modified. The difference in risk of Salmonella contamination in ground pork were then estimated between the baseline
and modified levels.
When the amount of deep tissue lymph nodes were changed from baseline value to zero, the probability of Salmonella
contaminated ground pork servings only reduced from 8.3% to 7.9%. This reflects a situation if the deep tissue lymph
nodes would be completely removed in processing plant. If the intervention of deep tissue lymph nodes is the one to
consider for the future to provide a high level of ground pork safety, the questions is, how likely is the intervention to
occur, and at what cost. In practice, the complete removal of deep tissue lymph nodes could be very time and labor
consuming. However, when the probability of Salmonella contamination on carcass surface at processing plant reduced
from baseline value to its half, the probability of Salmonella contaminated ground pork servings would reduce from 8.3%
to 6.6%, end even below 2.8% if carcass surfaces were completed decontaminated.
The findings reveal that deep tissue lymph nodes have non-significant impact, and Salmonella contamination from carcass
surface has a more important influence on Salmonella contamination in ground pork. Therefore, compared to intervention
strategies, such as mitigation of Salmonella on carcass surface, the intervention of DTLNs at processing plants might not
be able to effectively reduce the Salmonella contamination in ground pork.
(08-249) Expansion of the Salmonella in Pork Risk Assessment: Incorporation of Toxoplasma in pork,
and assessing the relationship between on-farm factors and risk.
We extended our previous farm-to-illness model (NPB project number 07-079) to incorporate the risk of human
toxoplasmosis as well as salmonellosis, from the consumption of both fresh intact pork cuts, and mixed pork as
represented by breakfast sausage. In addition, the effect of two different production methods: continuous flow and all-in/
all-out, is explored.
This risk assessment is based on a probabilistic simulation which models the variation that exists in the various factors
affecting the addition, growth, partitioning, and inactivation of pathogens along the pork production chain. The prevalence
and level of Salmonella contamination is simulated by the model, beginning with the arrival of weaned pigs at the
production site, and tracked through growing, slaughter, processing and consumer cooking to the point of consumption.
Toxoplasma level and prevalence in intact meat is represented by estimates from the literature, and the model uses
simulated and literature values as input to the mixed meat module, which predicts the level and prevalence of Salmonella
and Toxoplasma contamination in servings of breakfast sausage. The effect of cooking on both pathogens is simulated to
yield a dose per cooked serving, which is translated into a risk of illness by dose-response models. In the absence of an
accepted dose-response model for Toxoplasma cysts in humans, the exponential model is used with two different r-values.
The mean risk of salmonellosis per serving predicted by the model is 2×10-6 for intact cuts of pork and 6×10-6 for
breakfast sausage. This translates to 50,000 cases of illness annually from fresh intact pork, and 30,000 for breakfast
sausage, using available consumption data.
Depending on the r-value applied, the mean risk of toxoplasmosis predicted per serving of intact pork ranged from
8×10-7 to 8×10-6, and that per serving of breakfast sausage ranged from 7×10-7 to 7×10-6, leading to between 3,000 and
35,000 annual cases from consumption in the home of these two products.
The model supports scenario analysis to explore the impact of:
• proportion of pigs from all-in/all-out sites among pork production facilities
• prevalence of Toxoplasma among pigs contributing to mixed meat
• prevalence of Toxoplasma among grower pigs
• two levels of Toxoplasma cysts in pork
• efficacy of washing of viscera
• proportion of pathogens in protected areas during cooking
• probability of undercooking
This project is of interest to stakeholders at all stages of the pork production chain. The model developed facilitates
further understanding of opportunities to manage the risk of both salmonellosis and toxoplasmosis due to pork products.
It can also be used to identify important research opportunities.
(07-079) A Quantitative Risk Assessment for Salmonella spp. in Pork Meat

We developed a farm-to-illness risk assessment to investigate the risk of human salmonellosis from the consumption
of fresh pork. The risk assessment explores major issues in both pre-harvest and post-harvest pork safety including
on-farm transmission of Salmonella, the effect of various slaughter processes on both the prevalence and the level of
the pathogen, and the persistence of contamination through distribution, storage, and preparation. The model is a
probabilistic simulation capturing the real-world variation that is inherent in pork production systems. The risk assessment
tracks the prevalence and level of Salmonella contamination on carcasses through grow-out, multiple stages of slaughter,
retail and consumer storage and consumer cooking in the home environment. The result is a prediction of the level of
Salmonella ingested by an individual from the consumption of a serving of pork. The probability that this will result in
salmonellosis is finally estimated. Model results estimate that the risk of salmonellosis from the consumption of fresh pork
meat products prepared in the home is 8×10-7 (mean estimate), which translates to 0.8 illnesses per million servings.
Based upon available consumption data, we estimate this would result in 8,120 cases of salmonellosis per year in the
US. (Note that the scope of the work included only fresh pork meat products, and does not include RTE or mixed meat
products, or those prepared outside the home).
Since the specific values of the inputs on which the model is based (for example prevalence of Salmonella in weaners), can
be varied to explore the effect of different control measures, scenario analysis permits a comparison of these approaches
in terms of their ultimate impact on the risk of illness. Model analyses indicate that key parameters in the pork production
chain that have a strong influence upon the risk of illness are:
• the between herd and within herd prevalence at slaughter,
• the change in prevalence that occurs during lairage,
• the amount of gut contents added to a carcass during evisceration if rupture of the viscera occurs.
Of less importance in controlling the risk of illness are:
• The cross contamination that occurs during scald
• Probability of external contamination on pigs at slaughter that are not infected with Salmonella.
This project is of interest to stakeholders at all stages of the pork production chain.
(06-144) Risk Assessment for Campylobacter spp. and Salmonella spp. in Pork
The purpose of this research was to develop a retail-to-table risk assessment to evaluate the likelihood of developing
salmonellosis from consumption of a serving of enhanced or un-enhanced pork chops (bone-in) and roasts (boneless) cooked
to end point temperatures of 145–160º F. With guidance received from USDA-FSIS, that Salmonella is the target organism of
choice for cooking standards for fresh pork, and after NPB concurrence, Exponent did not consider Campylobacter in this risk
assessment. Salmonella exposure was estimated using a statistical method that repeatedly (10,000 times) picked data points
from distributions including: Salmonella levels in retail pork; Salmonella growth during home transport and storage; and
Salmonella levels after cooking and at consumption. The report describes all databases and the rationale and approach for
their use. Risk was characterized as risk of illness per serving. No Salmonella survived cooking to 160°F or to 145, 150 and 155°
F in 99% of the simulations. However, following USDA’s recommendation, a low illness risk was estimated at extreme abuse
conditions (high storage and transport temperature and long storage times) at temperatures below 160°F for some cuts.
Potential error and uncertainty sources of the model are described. Currently, USDA recommends consumers to cook pork
to160-170°F. However, FDA’s institutional and retail establishment guidance permits 145°F. This study shows that pork cooked
to 145 ºF for 15 seconds does not increase salmonellosis risk, if good pre-cooking handling practices are used. Further, a
greater safety margin is achieved for products cooked to Food Code recommendations.
(98-199) Salmonella Risk Assessment for Blade Tenderized, Immersion marinated, Needle Injected and
Fibrinogen Processed Pork Cuts During Processing, Storage and Cooking
The objectives of this extensive research project funded by NPPC were to evaluate commonly used technologies in
the pork processing industry to produce products of highest uniformity, sensory value, and yield for their effects on
microbiological safety. Once this was determined, various cooking protocols were evaluated to determine their adequacy
in eliminating levels of pathogenic contamination resulting from these processing technologies. Specifically, the
technologies studied were blade tenderization, needle injection, Fibrimex restructuring, and in bag rub marination.
CONTROL STRATEGIES
(18-100) Post-Harvest Hurdle Interventions to Reduce Salmonella in Pork Trimming

(18-049) Interventions to control Salmonella I 4, [5], 12 :i:- in Pork

(17-047) Efficacy of Interventions to reduce Salmonella on Fresh Pork Products to Improve Food Safety
(09-065) Quantifying the Effect of Slow-Cooking Operations on the Thermal Resistance of Salmonella in
Whole-Muscle Pork Products
The overall goal of this project was to improve the reliability of thermal process validation tools for Salmonella in pork
products, by accounting for stress adaptation that can occur during slow cooking processes. The specific objectives were:
(1) To modify, for ground and whole-muscle pork products, a model recently developed at MSU to predict the rate of
Salmonella thermal inactivation as a function of both product temperature and prior (sub-lethal) thermal history, and (2)
To validate this model via pilot-scale challenge studies using ground and whole-muscle pork products inoculated with
Salmonella. The project entailed thermal treatment (cooking) trials in four different systems, ranging from 1 g samples
in a highly controlled laboratory system to ground pork patties and whole-muscle chops and roasts in two pilot-scale
oven systems. The key results were: (1) The thermal resistance of Salmonella in pork products increases when subjected
to sublethal injury during slow cooking processes, and an improved inactivation model accounts for that effect; (2) The
thermal resistance of Salmonella is significantly greater (~50%) in whole-muscle pork than in ground pork; (3) Application
of laboratory-based inactivation data for commercial process validations carries inherent uncertainty that is generally
underreported and that increases with scale-up; and (4) Although slow cooking processes can allow Salmonella to adapt
and become more thermally resistant, moist-air cooking to an endpoint ≥160°F generally results in sufficient cumulative
lethality to overcome this effect. These findings mean that producers of ready-to-eat pork products need to ensure that
process validations are based on inactivation data and models that are appropriate to their specific product and process
(e.g., whole-muscle vs. ground) and that thermal processes are designed with a sufficient margin of safety to account for
the inherent uncertainty associated with the application of inactivation models to thermal process validations.
(08-171) Effect of lactic acid commercial chilling processes on the survival of Salmonella,
Campylobacter coli and Yersinia spp. in pork variety meats
Previous research on pork variety meats reported a 2% lactic acid spray was the most effective antimicrobial of those
tested with the least negative effect on product quality (Zerby et al., 1998). However, as there has been limited if any
implementation of these interventions, the purpose of the proposed research was to validate the effect of current
industry practices of chilling and freezing pork variety meats, with and without the application of lactic acid, on the
survival of pathogens of concern to the pork industry. Four different variety meats (liver, heart, intestines and stomach)
were inoculated with three different pathogens (Salmonella, Campylobacter and Yersinia) and subjected to four different
chilling treatments (with and without the application of 2% lactic acid). Samples were taken before treatment, after
treatment and throughout a six-month shelf-life to measure the level of reduction or survival of each organism. Additional
testing was performed on uninoculated samples to evaluate the reduction and survival of aerobic plate count, total
coliforms and Escherichia coli before and after treatment and throughout shelf-life. Although an extensive baseline was
developed and intervention strategies were evaluated for pork variety meats in 1998, the majority of the pork industry
continues to depend on conventional refrigeration and freezing as their primary methods for controlling microbial growth
in these products.
With the reported increase in export of pork products, U.S. pork and pork variety meats are now more than ever
before subject to the scrutiny and expectations of the countries we trade with. Some countries are more accepting
of antimicrobial interventions such as organic acids than other countries. While other countries may approve of
the application of water-based antimicrobial treatments, such as hot water, they may not approve of water-based
antimicrobials containing additional levels of chlorine that might serve as an effective antimicrobial. In addition, due to the
low relative monetary value of variety meats as compared to carcass meats, most decontamination interventions during
the slaughter process are directed at carcass meats. Therefore, a company may be totally dependent on refrigeration and/
or freezing for microbial control. Utilizing validated and published procedures, whether they are antimicrobial treatments
of a chemical nature or commercial refrigeration and freezing, will greatly support processors of pork variety meats
whether they do business on a domestic or an international level. Overall, the use of 2% lactic acid as a decontamination
intervention in addition to good GMPs (employee hygiene, sanitation, and rapid chilling) during processing of pork variety
meats results in significant reductions in levels of Salmonella, Y. enterocolitica, and C. coli, as well as indicator organisms
(APC, ECC, and TCC). However, significant reductions were also observed on variety meats treated with only a water wash
and subsequently frozen. Trends in levels of APC, ECC, and TCC are similar to trends in levels of pathogens and can thus
be used as surrogates for pathogens to monitor and validate processes in plants. Results did vary between the different
types of variety meats suggesting that one intervention may not be suitable for all products. The results of this study may
be used to support industry best practices for reducing pathogens in pork variety meats destined for export.
(00-102) Model for the reduction of Salmonella on swine carcasses in slaughter facilities. I. Location of
probable sources and sites of contamination
Carcass swabs, carcass cavity swabs, head meat and trim samples were analyzed for populations of coliforms, Escherichia
coli Biotype I, and the presence of Salmonellae. Approximately 560 samples were analyzed over a 12 month period. Only
5 samples (2 carcass cavity and three head meat) tested positive for the presence of Salmonellae. For all 5 samples, the
corresponding populations of either coliforms or Escherichia coli Biotype I were below a statistically calculated value of
“mean + 2 standard deviations”, and in 4 of the 5 samples, the populations of both indicator bacteria were actually below
the calculated mean for the sample set. There were insufficient Salmonellae positive samples to draw any statistical valid
conclusions. However, populations of neither coliforms nor Escherichia coli Biotype I were predictors of the presence or
absence of Salmonellae.
(97-1895) Reduction of Salmonella Contamination on Pork and Shelf Life Extension Using Radiant Wall
Oven Heating
Microbial contamination with foodborne pathogens continues to be a serious problem for the pork industry. The heightened
media coverage of the meat industry has brought the concern of food safety to the attention of consumers. In response, the
meat industry is aggressively pursuing methods to reduce the level of foodborne pathogens on its product. Developments
that are practical and effective in pathogen reduction can aid the industry and enhance product safety.
This study revealed that initial efforts using the radiant wall oven has some potential in addressing the issue of pathogen
reduction. The radiant wall oven is relatively new and allows a product to be subjected to radiant heat to temperatures
up to 1500°C for just a few seconds. From a microbiological standpoint, this thermal treatment can reduce the level
of bacterial contamination on the food surface while resulting in minimal change in the food itself. By subjecting pork
loins to 1200-1300°F for 3 seconds, populations of Salmonella (including S. Typhimurium DT 104 and the heat resistant S.
seftenburg) were reduced by 1-1.5 logs while the product retained a fresh-like appearance.
Section 4:
Environmental Impact
(13-260) Systematic Review of Antibiotic Resistance of Public Health Importance in Environments near
Swine Operations
Systematic reviews are a rigorous knowledge synthesis technique first developed in the health sciences to summarize
information from numerous randomized trials examining the clinical efficacy of an intervention. In contrast to a narrative
review, this method permits the evaluation of all available evidence on the question of interest using a standardized
process. Because of varied opinions regarding the evidence for sources of antibiotic resistance found in the environment,
this approach is ideally suited to this type of topic. We conducted a systematic review to address the following question:
Is the prevalence or concentration of antibiotic resistant bacteria (ARB) or resistance genes (ARG) in soil, water, air or
free-living wildlife higher in close proximity to, or downstream from, known or suspected sources compared to areas more
distant from or upstream from these sources? The total number of studies that were evaluated was 4,524, and from these,
60 were included in the final evaluation (19 with ARG outcome and 41 with ARB outcome). Many of the studies provided
little useful information due to a high level of bias in the study design. The systematic review found a large number
of studies presenting qualitative evidence that proximity to or direction from point sources, particularly waste water
treatment plants, may be associated with higher levels of antibiotic resistance; however, very few studies quantitatively
characterize this effect or provide statistical inference to aid in interpretation. This systematic review provides a strong
imperative to improve research methods in order to provide interpretable, quantitative information about the effect of
point sources on resistance in the environment.
(13-006) Evaluate the dissemination of Salmonella in the environment following land application of
swine manure
There is tremendous pressure on the US pork industry to ban the prophylactic and growth promotion use of antimicrobials
in feed due to the generation of antimicrobial resistant (AMR) bacterial strains. An important concern is the dissemination
of AMR Salmonella in the environment after swine manure application. The main objective of this study was to determine
the potential role of lagoons and manure pits in the transmission of AMR Salmonella in the environment following land
application of swine manure on commercial swine farms in Iowa (n=7) and North Carolina (n=6). In Iowa the manure is
stored in pits and applied using an injection system while manure in NC is stored in lagoons and applied directly on
the soil using a spray method. Salmonella prevalence was compared on these conventional swine farms at different
time points including day 0 (before and after manure application) and subsequently on days 7, 14 and 28 post manure
application on specific land locations at every site. The samples consisted of lagoon/manure pit and soil on day 0 while
only soil samples were collected on the following sampling time points. Overall, we collected a total of 1,200 soil samples
(IA=700; NC=500) and 50 lagoon and 70 manure pit samples from NC and IA, respectively. Overall Salmonella prevalence
was 13.33% (176/1320) while the prevalence in soil and lagoon were 10.92% and 37.5%, respectively. The Salmonella
prevalence in North Carolina (28.18%) was significantly higher than in Iowa (2.73%) (p < 0.001). We detected a significant
decrease in prevalence of Salmonella from the marked areas as we moved from Day 0 to Day 21. We identified 12
serotypes, however, it is important to note that no serotype found in one state was detected from the other highlighting
serotype association based on geographic region. For example, we detected Anatum (7.39%), Litchffield (3.98%), and
Infantis (0.57%) in IA, while Altona (7.95%), Derby (3.98%), Johannesburg (3.98%), Mbandaka (1.70%), Muenster (9.09%),
Rissen (0.57%), Typhimurium var5- (20.45%), Uganda (2.27%), and Worthington (5.68%) in NC. A total of 80.47% of the
Salmonella isolates were multidrug resistant (MDR; resistance to three or more antimicrobials) with the most frequent
AMR against Streptomycin (82.81%), sulfisoxazole (73.44%), and kanamycin (61.72%). PFGE genotyping revealed genotype
relatedness among Salmonella recovered from lagoon and soil at multiple time points with relatively close geographic
proximity and serotypes. Our study highlights Salmonella transmission in the environment in commercial swine farms
is dependent on the type of manure storage and its application method. Finally, the rapid decline in the prevalence of
Salmonella in soil samples on subsequent collection days (Days 7, 14, 21) clearly indicates the inability of this pathogen to
survive in the environment for longer durations.
(05-041) Longitudinal evaluation of the effect of ventilation and environmental management of swine
barns on Salmonella prevalence in finishing swine
These data suggest that environmental temperature; in particular lower environmental temperatures in early finisher pigs
and high environmental temperature in older finisher pigs are associated with increased Salmonella prevalence. This holds
promise for designing interventions pre-harvest that have positive impacts on swine health, production performance
and food safety. Further studies to understand the impact of environmental management of swine barns under different
building designs and different geographical regions are needed. Furthermore, more intensive monitoring of both the
barn environment and Salmonella shedding is necessary to understand the time period between suboptimal thermal
conditions and Salmonella shedding in swine.
(99-112) Reduction of Salmonella and Other Fecal Microbes in Swine Waste Treatment Systems
The presence and control of human pathogens in waste from commercial swine farms has emerged as a public health
and policy issue that impacts management practices for swine waste treatment. This project focused on quantifying the
reductions of fecal microbes in commercial swine waste lagoons and constructed wetlands. Constructed wetlands are a
promising alternative or additional treatment technique for flushed swine waste. Swine waste samples were analyzed for
a suite of six microbial indicators (fecal coliforms, Escherichia coli, enterococci, Clostridium perfringens spores, somatic
coliphages and male-specific coliphages) and Salmonella spp., a group of pathogenic bacteria. In untreated swine waste
from flushed and pit-plug systems at four swine farms, the average concentration of Salmonella was measured to be
3800 MPN/100 mL. Salmonella were reduced by approximately 96% in primary anaerobic lagoons at these farms and by
a further 97% in the secondary lagoons used at two of the farms. In general, fecal coliforms, E. coli and enterococci were
reduced to a similar, but slightly greater, extent than Salmonella (» 97-98%) in each lagoon cell provided. C. perfringens
spores, investigated as a potential model for the removal of helminth ova and protozoan parasite cysts and oocysts in
swine waste, were less efficiently reduced in lagoons than the other enteric bacteria studied, being reduced by an average
of 84% in primary lagoons and another 92% in secondary lagoons. Somatic coliphages and F+ coliphages, investigated as
potential models for the removal of enteric viruses pathogenic to humans, were reduced to a similar extent as measured
for fecal coliforms, E. coli and enterococci (» 97% in primary lagoons and a further 96% in secondary lagoons). In a field-
scale surface flow (SF) constructed wetland operated as a secondary treatment system receiving anaerobic lagoon liquid,
fecal coliforms, E. coli and enterococci were reduced by 98, 99 and 87%, respectively. Salmonella were reduced by 96% in
this constructed wetlands system, and C. perfringens spores, somatic coliphages and male-specific coliphages by 97, 99,
and 98%, respectively. In laboratory-scale SF and subsurface flow (SSF) wetland reactors, temperature and loading rate
were shown to be significant variables affecting the performance of the reactors for reducing concentrations of enteric
microbes and nutrients in swine lagoon liquid. At temperatures of 10, 20 and 30°C and total Kjeldahl nitrogen loading
rates of 10, 25 and 40 kg/ha/d, the vegetated SSF reactor generally achieved higher microbial and nutrient reductions
than either the vegetated SF reactor or the unvegetated SSF control reactor. The results of this study show that significant
reductions of fecal microbes can occur in anaerobic swine waste lagoons, but that high concentrations of these microbes
remain in lagoon liquid. Alternative, or additional, treatment using constructed wetlands can achieve significant removal
of pathogens and nutrients from swine waste if important design variables such as temperature and mass/hydraulic
loading rates are considered.
Notes
Notes
Salmonella
RESEARCH COMPENDIUM
National Pork Board

1776 NW 114th St
Clive, IA 50325
pork.org • (800) 456-7675
©2018 National Pork Board, Des Moines, IA USA. This message funded by America’s Pork Producers and the Pork Checkoff. #04939 12/18

Salmonella

Uploaded by

Copyright:

Available Formats

You might also like

Salmonella

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Salmonella

Uploaded by

Copyright:

Available Formats

Salmonella

pork.org • (800) 456-7675

Systems perspective on Salmonella sources in pork production and processing.

Additional Salmonella and Pork Safety Resources

Section 1: Pre-Harvest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-25

• Pathogenesis & isolates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

• Diagnostic tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

• On-farm prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15

• Control strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19

• Antibiotic resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22

Section 2: Peri-Harvest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26-31

• Control strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29

Section 3: Post-Harvest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32-43

• Prevalence in pork and pork products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32

• Controlling Multi-Drug Resistant (MDR) bacteria in pork . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36

• Diagnostic tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37

• Risk assessments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39

• Control strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42

Section 4: Environmental Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44-45

All full research projects final reports are available at pork.org/research.

(12-069) Salmonella serovar distribution and persistence in finisher sites

(09-120) Molecular Basis of Salmonella Competition in Broth Culture

(98-194) Colonization Pattern of Salmonella Typhimurium DT104

(09-094) Development and validation of molecular-based tools to differentiate attenuated Salmonella

(06-185) Salmonella sample cryopreservation enables epidemiologic study, quantification of bacteria

(01-063) Validation of a Diagnostic Test to Identify Salmonella enterica in Swine: Preliminary

(99-230) Reduction of Salmonella by Bacteriophage Treatment

(99-136) A Rapid, Specific Test for Salmonella Subtypes

(15-070) An analysis of primary research and meta-analysis to investigate temporal changes in

(14-288) Assessment of International Swine Programs for On-farm Salmonella Control

(04-100) Salmonella prevalence during the finishing production stage

(00-100) Evaluation of Methods to Differentiate Herd Responses to Salmonella

(98-148) Salmonella Abatement in the Pork Chain

(05-193) Use of naturally-occurring bacteriophage to reduce Salmonella in swine prior to harvest

(01-077) Lactobacillus acidophilus as Probiotic to Control Salmonella in Swine

(15-072) Effects of antibiotics on intestinal and extra-intestinal multidrug-resistant and pan-susceptible

(98-218) Characterization of Multiple-Antimicrobial Resistance among Swine Salmonella

(06-054) Quantifying Salmonella and Campylobacter Peri-harvest in Swine

(00-098) Evaluation of Antemortem Salmonella Detection Procedures in Market Weight Swine

(97-2002) Affect of Feed Withdrawal on the Shedding of Salmonella Typhimurium by Swine

(97-1900) The Influence of Competitive Exclusion Cultures and Fructooligosaccharide Supplementation

(00-136) Effect of Sodium Chlorate on Salmonella Typhimurium in the Pig Gut

(12-145) National Prevalence of Salmonella Contamination in Retail Ground Pork

(10-146) Evaluating the sources of Salmonella after carcass chilling

(07-050) Enumeration of Salmonella throughout the Pork Harvesting Process

(08-120) Comparative Evaluation of Rapid Methods for Salmonella Detection in Pork

(02-027) The Development of a Novel Immunosensor to Detect Salmonella

(10-130) Risk informed management of Salmonella in deep tissue lymph nodes

(07-079) A Quantitative Risk Assessment for Salmonella spp. in Pork Meat

(18-100) Post-Harvest Hurdle Interventions to Reduce Salmonella in Pork Trimming

(18-049) Interventions to control Salmonella I 4, [5], 12 :i:- in Pork

National Pork Board

You might also like