APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 2008, p. 5130–5138
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 74, No. 16
Molecular Epidemiology of Campylobacter jejuni Populations in Dairy
Cattle, Wildlife, and the Environment in a Farmland Area?†
Patrick S. L. Kwan,1Mishele Barrigas,6Frederick J. Bolton,2Nigel P. French,5Peter Gowland,6
Richard Kemp,3Howard Leatherbarrow,3Mathew Upton,1and Andrew J. Fox2,4*
Department of Medical Microbiology, School of Medicine, University of Manchester, Clinical Sciences Building, Manchester Royal Infirmary,
Manchester M13 9WL, United Kingdom1; Health Protection Agency North West Laboratory, Clinical Sciences Building,
Manchester Royal Infirmary, Manchester M13 9WZ, United Kingdom2; Epidemiology Group, Faculty of Veterinary Science,
University of Liverpool, Neston CH64 7TE, United Kingdom3; FEMS North West, Royal Preston Hospital, Sharoe Green Lane,
Preston PR2 9HG, United Kingdom4; Institute of Veterinary, Animal and Biomedical Sciences, College of Sciences,
Massey University, Palmerston North, New Zealand5; and Biological Sciences, University of
Staffordshire, Stoke-on-Trent, United Kingdom6
Received 26 September 2007/Accepted 8 June 2008
We describe a cross-sectional study of the molecular epidemiology of Campylobacter jejuni in a dairy farmland
environment, with the aim of elucidating the dynamics of horizontal transmission of C. jejuni genotypes among
area of farmland in northwest England was characterized by multilocus sequence typing. A total of 91 sequence
types and 18 clonal complexes were identified. Clonal complexes ST-21, ST-45, and ST-61, which have been
frequently associated with human disease, were the most commonly recovered genotypes in this study. In addition,
widely distributed genotypes as well as potentially host-associated genotypes have been identified, which suggests
that both restricted and interconnecting pathways of transmission may be operating in the dairy farmland envi-
ronment. In particular, the ST-61 complex and the ST-21 complex were significantly associated with cattle. In
contrast, complex strains ST-45, ST-952, and ST-677 were isolated predominantly from wild birds, wild rabbits, and
environmental water. A considerable number of novel sequence types have also been identified, which were
unassigned to existing clonal complexes and were frequently isolated from wildlife and environmental sources. The
segregated distribution of genotypes among samples from different sources suggests that their transmission to
humans is perhaps via independent routes. Insight into the dynamics and interactions of C. jejuni populations
between important animal reservoirs and their surrounding environment would improve the identification of
sources of Campylobacter infection and the design of control strategies.
Campylobacteriosis is the most frequent explanation of
acute bacterial gastroenteritis and is a zoonotically transmitted
disease caused primarily by the pathogen Campylobacter jejuni
via food-borne routes. Campylobacter infections have a sub-
stantial public health and socioeconomic impact (28, 31); over
46,000 cases have been reported in England and Wales in 2006
(http://www.hpa.org.uk/infections), and an estimated 2 million
cases and 2,000 deaths in the United States are attributed to
Campylobacter infections annually (33).
C. jejuni has a widespread distribution with a broad range of
animal hosts and environmental reservoirs. Food-producing
animals such as poultry (6), cattle, and sheep (22, 32) com-
monly harbor vast numbers of C. jejuni isolates in their gas-
trointestinal tracts and represent an important route through
which organisms could enter the food chain. In addition, wild-
life such as wild birds (2, 3, 34) and mammals (27) are also
common hosts of the organism. C. jejuni isolates are also found
in the environment, including in surface waters from rivers,
streams, ponds, and agricultural runoffs (15, 17, 18) and in sand
from bathing beaches (1), and have been implicated in out-
breaks of campylobacteriosis (13, 26). According to a recent
survey, Campylobacter spp. were implicated in 14% of water-
borne infectious intestinal disease outbreaks from 1992 to 2003
in England and Wales, where the common sources were
springs, wells, and boreholes (30).
The transmission dynamics of this sporadic disease are com-
plex, and the relative contribution of different sources in C.
jejuni infections has remained unclear, reflected in part by a
lack of reliable and definitive characterization methods. Recently,
multilocus sequence typing (MLST) has been proven to be a
and has provided useful information in epidemiological and pop-
ulation dynamic studies of this important pathogen.
Following the recent MLST studies that have highlighted the
importance of cattle in C. jejuni epidemiology (12, 22), we
further present a cross-sectional study that is the final part of
a detailed survey of C. jejuni isolates in a 100-km2dairy farm-
ing area in Cheshire, United Kingdom (12), with the primary
objective of detailing the molecular epidemiology of C. jejuni
in dairy cattle and the surrounding farming environment in
order to elucidate the significance of horizontal transmission
among different sources. The distribution and diversity of C.
jejuni genotypes among cattle, wildlife, and environmental
sources were investigated from a population snapshot based on
* Corresponding author. Mailing address: Molecular Epidemiology,
Manchester Medical Microbiology Partnership, P.O. Box 209, Clinical
Sciences Building, Manchester Royal Infirmary, Oxford Road,
Manchester M13 9WZ, United Kingdom. Phone: 44 161 276 5689. Fax:
44 161 276 5744. E-mail: email@example.com.
† Supplemental material for this article may be found at http://aem
?Published ahead of print on 27 June 2008.
the systematic sampling of isolates within the area. Isolates
included in the study originated from a wide range of sources,
including cattle, sheep, wild birds and ducks, rabbits, badgers,
and environmental sources such as groundwater, soil, and a
MATERIALS AND METHODS
Study site and sample collection. A total of 100 km2of rural dairy farming area
in Cheshire, United Kingdom was systematically sampled for cattle and wildlife
feces and for soil and water, where the regional geographical features, study
design, and protocol for spatially structured sampling have been detailed previ-
ously (4, 12, 23). Of the C. jejuni isolates investigated in this study (n ? 327), 247
were sampled between May and July 2000 from all sources, where the spatial
analysis and modeling for 172 isolates have been detailed previously (12). In
2001, an additional 80 isolates were obtained from cattle by sampling from fresh
fecal pats in the same locations that had been sampled in the year 2000 (see
Table S1 in the supplemental material).
Bacterial growth conditions and species differentiation. Fecal samples from
cattle and wildlife (1 ml), soil samples, and water-sample filters (0.2-?m pores;
Nalge Nunc International, Rochester, NY) were added to 9 ml of Campylobacter
enrichment broth (Lab M, Bury, United Kingdom) and incubated at 42°C for
24 h under microaerophilic conditions. The broth was inoculated onto Campy-
lobacter blood-free agar (modified charcoal cefoperazone deoxycholate agar)
containing an antibiotic supplement (cefoperazone and amphotericin; catalogue
no. X112 and X212; Lab M) and was further incubated for 48 h under the same
conditions. Three to four colonies of bacterial growth were subcultured onto
Columbia blood agar (Lab M) and incubated for 24 to 48 h. On the basis of
growth and colony morphology on agar under the described conditions, oxidase
and catalase tests, Gram staining, and cell shape and size, Campylobacter spp.
were presumptively identified and were stored frozen at ?70°C in a cryogenic
preservative (Microbank; Pro-Lab Diagnostics, Neston, United Kingdom) until
required for culture. Bacterial cells from subsequent culture were heated and
lysed to release DNA for PCR. Campylobacter species were identified by per-
forming single-reaction PCRs with previously described primers and conditions
MLST. Characterization of isolates by MLST was performed as per the pub-
lished scheme for C. jejuni isolates (10). Briefly, fragments of seven gene targets
for each isolate were amplified using the following published primers and reac-
tion conditions: aspA (aspartase A), glnA (glutamine synthase), gltA (citrate
synthase), glyA (serine hydroxymethyltransferase), pgm (phosphoglucomutase),
tkt (transketolase), and uncA (ATP synthase alpha subunit). Amplicons were
purified with the MultiScreen PCR filter plate and MultiScreen vacuum manifold
(Millipore Corporation) according to the manufacturer’s instructions. Dideoxy
termination sequencing reactions were performed at least once on each of the
amplified forward and reverse DNA strands in 10-?l volumes containing a
10-?M primer, BigDye reaction mix version 3.1, and 5? sequencing buffer (PE
Applied Biosystems). The reaction conditions were 30 cycles at 96°C for 20 s,
50°C for 20 s, and 60°C for 4 min. Sequencher software version 4.0 (Gene Codes
Corporation) was used for sequence editing and assembly on a computer. The
Campylobacter MLST database (http://pubmlst.org/campylobacter) (16) was
used for sequence typing and the clonal complex assignment of C. jejuni isolates.
Statistical and gene flow analysis. The association between C. jejuni isolate
genotypes and isolation sources was analyzed with statistical tests using SPSS
software version 14.0, where pairwise comparisons of clonal complex distribu-
tions between isolation sources using the Yates’ continuity-corrected ?2test or
Fisher’s exact two-tailed test were performed. P values of ?0.05 were the criteria
chosen for statistical significance. Isolation sources and clonal complexes with a
small number of isolates in the data set were not included.
Estimates of the gene flow between C. jejuni populations from different
sources were carried out using DnaSP software (29). Gene flow is measured by
an FSTvalue that falls between 0 and 1, where a value of 0 indicates that the two
populations are indistinguishable, and a value of 1 indicates that the two popu-
lations are genetically distinct.
C. jejuni genotypes in the farmland environment. C. jejuni
isolates (n ? 327) from different animal and environmental
sources within the dairy farmland sampling area were charac-
terized by MLST. Among the 91 sequence types identified, the
most prevalent sequence types were ST-45 (n ? 44; 13.5%),
ST-21 (n ? 35; 10.7%), and ST-61 (n ? 31; 9.5%), which
collectively represented 33.7% of the entire data set, while
16.8% of the data set was represented by 55 sequence types
which appeared only once. Fifty-nine sequence types (n ? 283;
86.5%) were grouped into 18 clonal complexes, while the re-
maining 32 sequence types (n ? 44; 13.5%) were unassigned to
any clonal complexes (Fig. 1; Table 1).
Sixty percent of the data set belonged to three predominant
clonal complexes, namely, the ST-21 complex (n ? 77; 23.5%),
the ST-45 complex (n ? 71; 21.7%), and the ST-61 complex (n ?
49; 15.0%). Each of the 15 remaining clonal complexes was
represented by 6.7% or fewer isolates in the data set. Nine
clonal complexes were represented by just one sequence type,
five of which occurred only once. From 25 isolates, 21 sequence
types (ST-848, ST-851 to ST-853, ST-864 to ST-865, ST-944 to
ST-957, and ST-959, inclusive) were newly described in this
Ruminant isolates. A total of 162 C. jejuni isolates from
cattle were analyzed in this study. Forty-nine sequence types
were identified, from which 30 sequence types appeared only
once. The most prevalent sequence types were ST-61 (n ? 27;
16.7%) and ST-21 (n ? 24; 14.8%), and each of the 47 remain-
ing sequence types was represented by up to 6.2% of isolates in
the cattle data set (Table 1). All isolates were grouped into 13
clonal complexes, except for five unassigned sequence types
(3.1%), three of which were described for the first time in this
study (ST-851, ST-853, and ST-865).
The majority of the cattle data set (n ? 139; 85.8%) be-
longed to five clonal complexes (complexes ST-21, ST-61, ST-
48, ST-42, and ST-45), and 60% (n ? 97) belonged to either
the ST-21 complex or the ST-61 complex (Fig. 2). Each of the
eight remaining clonal complexes was represented by 3.7% or
fewer cattle isolates. Six clonal complexes were represented by
only one sequence type, from which five complexes appeared
only once in the data set.
The cattle data set was derived from samples obtained 12
months apart, designated group A (n ? 82; collected in 2000)
and group B (n ? 80; collected in 2001) (Table 1). Conse-
quently, 34 sequence types and 13 clonal complexes were iden-
tified in group A, while 29 sequence types and 8 clonal com-
plexes were identified in group B. A number of genotypes
overlapped between the two groups; 14 sequence types were
FIG. 1. Occurrence of clonal complexes of 327 C. jejuni isolates
from within a farmland environment. UA, unassigned sequence types.
VOL. 74, 2008MOLECULAR EPIDEMIOLOGY OF C. JEJUNI5131
TABLE 1. Distribution of C. jejuni among sources from 327 isolates
% of cattle in: % of isolates from:No. of isolates from:
RabbitBird WaterSoilTrough Badger Sheep Duck
Total 39.0 30.034.623.8 6.3 5.7 77 23.5
Total11.0 5.0 8.023.87121.7
Total22.028.8 25.39.5 3.12.9 4915.0
Total7.3 8.88.0 7.1 4.72.9 226.7
Total 6.113.8 9.95.7 185.5
Total1.2 0.69.56.38.6 123.7
Continued on following page
5132KWAN ET AL.APPL. ENVIRON. MICROBIOL.
% of cattle in:% of isolates from: No. of isolates from:
RabbitBird Water Soil TroughBadger Sheep Duck
Total 4.8 3.1 11.48 2.4
Total 3.73.8 3.76 1.8
ST-257ST-257 2.4 2.52.54 1.2
Total1.22.5 1.93 0.9
ST-49 ST-491.2 0.6a
ST-22 ST-221.2 0.6a
ST-283 ST-26711 0.3
Total 1.2 5.03.1 19.035.917.12 4413.5
aSequence types represented by a single isolate within a source.
bUA, unassigned sequence types.
VOL. 74, 2008 MOLECULAR EPIDEMIOLOGY OF C. JEJUNI 5133
shared, and all clonal complexes identified in group B also
appeared in group A (Fig. 3). Sequence types and clonal com-
plexes which did not overlap were represented by no more than
two isolates in each group. In addition, C. jejuni clonal com-
plexes (complexes ST-21, ST-61, ST-48, ST-42, and ST-45)
were predominant in both groups and collectively represented
85.4% and 86.4% of isolates in group A and group B, respec-
tively. Furthermore, there appeared to be a comparable prev-
alence between group A and group B for clonal complexes
ST-21 (39.0% and 30.0%, respectively), ST-61 (22.0% and
28.8%, respectively), and ST-42 (7.3% and 8.8%, respectively).
Only five isolates from sheep were characterized, with three
isolates belonging to the ST-21 complex and two belonging to
the ST-45 complex.
Wildlife isolates. From the wild bird isolates (n ? 64), 31
sequence types were identified, of which 12 appeared only
once. No obvious dominant sequence type was observed, and
the most commonly encountered sequence type, ST-45, was
represented by only six isolates (9.4%) (Table 1). Nineteen
sequence types (n ? 41; 64.1%) were grouped into 10 clonal
complexes, and 12 sequence types (n ? 23; 35.9%) in the data
set were unassigned to clonal complexes.
The ST-45 complex was predominant in the wild bird data
set, represented by 20 isolates (31.3%), while all other nine
lineages were individually represented by 6.3% or fewer wild
bird isolates (Fig. 2). Five clonal complexes were represented
by only one sequence type, and two appeared only once in the
data set. In addition, two isolates were collected from ducks,
both of which were of the ST-21 genotype.
Among the 21 sequence types identified from rabbit isolates
(n ? 42), ST-45 was the most frequently identified sequence
type (n ? 8; 19.0%), while 8 other sequence types were rep-
resented by two to four (4.8% to 9.5%) isolates, and 12 se-
quence types appeared only once in the rabbit data set (Table
1). Thirteen sequence types (n ? 34, 81.0%) in the rabbit data
set were grouped into seven clonal complexes, and 8 sequence
types (n ? 8; 19.0%) were not assigned to clonal complexes.
The ST-45 complex and the ST-21 complex were equally dom-
inant and were represented by 10 isolates each (23.5%) (Fig.
2). All other clonal complexes were individually represented by
four or fewer rabbit isolates. Three clonal complexes were
represented by one sequence type, one of which was repre-
sented by a single isolate.
Of the 11 isolates obtained from badgers, the majority be-
longed to the ST-45 complex (n ? 9), with the remaining two
isolates belonging to clonal complexes ST-42 and ST-283.
Environmental isolates. Sixteen sequence types were iden-
tified from 35 C. jejuni isolates obtained from environmental
water samples, with 9 sequence types having occurred only
once (Table 1). ST-45 was the dominant sequence type (n ?
13; 37.1%), with all remaining sequence types being repre-
sented by three or fewer isolates.
Eleven sequence types (n ? 29; 82.9%) were grouped into
eight clonal complexes, and 6 sequence types (n ? 6, 17.1%)
were unassigned to clonal complexes. The ST-45 clonal com-
plex was the predominant genotype (42.9%), while the ST-957
complex (11.4%) and the ST-677 complex (8.6%) were the
next most common (Fig. 2). The five remaining clonal com-
FIG. 2. Distribution of clonal complexes of 303 C. jejuni isolates recovered from cattle, rabbits, wild birds, and environmental waters. UA,
unassigned sequence types.
FIG. 3. Distribution of C. jejuni clonal complexes among two
groups of cattle isolates from the same geographical area collected 12
months apart. UA, unassigned sequence types.
5134KWAN ET AL.APPL. ENVIRON. MICROBIOL.
plexes were represented by two or fewer isolates. Three clonal
complexes were represented by one sequence type, all of which
were represented by a single isolate.
Isolates obtained from soil samples (n ? 5) belonged to the
ST-45 complex (n ? 2) and the ST-42 complex (n ? 1), and two
sequence types (ST-838 and ST-950) were not assigned to a
clonal complex. A single isolate that originated from a drink-
ing-water trough belonged to the ST-61 complex.
Association between genotypes and sources. The frequency
distribution of C. jejuni clonal complexes described appeared
to be nonrandom among sources. Consequently, pairwise com-
parisons of the clonal complex distribution between isolation
sources using the ?2tests were performed to investigate pos-
sible associations (Table 2). However, certain isolation sources
and clonal complexes with a small number of isolates in the
data set were not included. Accordingly, isolates which origi-
nated from cattle, wild birds, wild rabbits, and environmental
sources (water, soil, and water trough, collectively), isolates
which belonged to clonal complexes ST-21, ST-45, ST-61, ST-
42, ST-48, ST-677, and ST-952, and those with unassigned
sequence types were analyzed.
The proportion of isolates which belonged to the ST-61
complex was found to be significantly higher in cattle than in
wild birds (P ? 0.0005), wild rabbits (P ? 0.047), and environ-
mental sources (P ? 0.008), but there were no significant
differences in the prevalence of the ST-61 complex within these
In contrast, the prevalence of the ST-45 complex was signif-
icantly higher in wild birds (P ? 0.0005), rabbits (P ? 0.011),
and environmental sources (P ? 0.0005) than in cattle, but
there were no significant differences in the prevalence of the
ST-45 complex among these sources.
A significantly higher prevalence of the ST-21 complex was
observed in both cattle and rabbits than in wild birds (P ?
0.0005 and P ? 0.020, respectively) and environmental isolates
(P ? 0.0005 and P ? 0.032, respectively), but no significant
difference in the prevalence of the ST-21 complex was ob-
served between cattle and rabbits.
The prevalence of the ST-42 complex was less variable
among sources, where no significant differences were found.
The presence of the ST-48 complex was detected only in cattle
and environmental sources, and the difference was significant
between cattle and wild bird (P ? 0.007) and rabbit popula-
tions (P ? 0.047).
The ST-677 complex was significantly associated with wild
birds (P ? 0.024), rabbits (P ? 0.007), and environmental
sources (P ? 0.028), but no significant differences were found
among these sources.
Similarly, the ST-952 complex was absent in cattle, signifi-
cantly associated with rabbits (P ? 0.042) and environmental
sources (P ? 0.001), and marginally associated with wild birds
(P ? 0.079), but no significant differences were found between
The proportion of isolates with unassigned sequence types
was also significantly higher in wild birds (P ? 0.0005), rabbits
(P ? 0.001), and environmental sources (P ? 0.001) than in
cattle, although there were no significant differences among
Gene flow analysis between isolation sources. Based on the
concatenated sequences of all seven housekeeping loci of C.
jejuni isolates from cattle, wild birds, wild rabbits, and the
environment, a gene flow analysis between these sources was
The FSTvalues generated from six pairwise comparisons
between isolate sources ranged from 0.009 to 0.199, reflecting
the presence of both genetically similar and moderately differ-
ent C. jejuni isolate populations (Table 3). In particular, higher
FSTvalues were observed for isolates from cattle and birds
(0.199) and cattle and environmental sources (0.175), indicat-
ing limited genetic exchange between these C. jejuni isolate
populations. In contrast, lower FSTvalues for the remaining
four comparisons, ranging from 0.009 between wild birds and
the environment to 0.083 between cattle and rabbits, suggest
that these C. jejuni isolates represent a genetically indistin-
A better understanding of the epidemiology of C. jejuni is
clearly necessary as it continues to be the most common bac-
terial cause of human gastroenteritis. Indeed, the routes of
disease transmission and the relative significance of infection
sources to the burden on human disease have largely remained
elusive. Nonetheless, several previous studies have greatly im-
proved the understanding of C. jejuni epidemiology by inves-
tigating C. jejuni isolate populations from a range of animal
hosts using geographically and temporally diverse isolate col-
lections which were characterized by MLST (5, 9, 22, 25).
TABLE 2. Summary of P values of significance in the pairwise comparison for the differences in the distribution of C. jejuni clonal complexes
between isolation sources
P value for cattle with indicated ST
P value for wild bird
P value for rabbit
TABLE 3. Pairwise FSTvalues of the gene flow analysis of C. jejuni
isolates between sources
FSTvalue for isolates from:
Cattle Wild birdRabbit
VOL. 74, 2008 MOLECULAR EPIDEMIOLOGY OF C. JEJUNI 5135
However, the dynamics and interactions of C. jejuni isolate
populations between animal reservoirs and their immediate
environment have not been established. Given the important
role of cattle in C. jejuni epidemiology (22, 32), this study
addresses the molecular epidemiology of C. jejuni isolate pop-
ulations in a dairy farmland environment, with the aim of
elucidating the dynamics of horizontal transmission of C. jejuni
genotypes among different sources within a defined area, spe-
cifically whether cattle would acquire C. jejuni isolates from the
environment and vice versa. Insight into such interactions
would advance the knowledge of C. jejuni epidemiology to
inform on the relative importance of potential reservoirs for
This study has demonstrated that clonal complexes ST-21,
ST-45, and ST-61 were the most common C. jejuni genotypes in
the dairy farming environment under study, comprising 60% of
the entire data set (Table 1; Fig. 1). This finding may have
significant implications for disease control and prevention, as
not only do these strains have the capacity to cause disease,
they are also the most frequently isolated genotypes in humans
(9). Moreover, seven further clonal complexes identified have
also been associated with human infections, albeit on a smaller
scale (9). These observations clearly highlight the need to rec-
ognize that cattle and their associated environment could act
as important reservoirs for human disease.
The genotypic composition of the C. jejuni population in
cattle was found to be consistent with that delineated from a
sizeable number of isolates in a previous longitudinal study
(22), where nine clonal complexes overlapped between studies.
Clonal complexes ST-21 and ST-61 were highly prevalent, and
the ST-48 complex was moderately common in both cases,
although a relatively lower number of ST-42 complex isolates
and higher number of ST-45 complex isolates were observed in
this study, whereas strains which did not overlap accounted for
less than 4% of the isolates in each data set. These findings
therefore confirmed the major C. jejuni genotypes found in
cattle from the longitudinal study; they further reinforce the
hypothesis that ST-61 complex isolates may be from a cattle-
adapted C. jejuni lineage, especially in the United Kingdom, as
suggested in previous studies (5, 9, 12, 22, 25). Further, since
all but one of the clonal complexes (ST-1332) identified in
cattle in this study have been associated with human infections
in the past, there is evidence to suggest that cattle may serve as
a major reservoir for C. jejuni infections. A longitudinal aspect
was also included in the cattle isolates in this study, where
isolates obtained on two occasions separated by 12 months had
similar genotype distributions (Fig. 3). This indicates that C.
jejuni genotypes in dairy cattle were largely stable over time,
which would concur with the finding that there were minimal
seasonal patterns in C. jejuni genotypes in cattle from the
previous longitudinal study (22).
The ST-45 complex was found to be the most prevalent
(31.3%) and the only dominant genotype among wild birds in
this study, which was apparent by comparing the ST-45 com-
plex prevalence to that of the next most common lineages, the
ST-21 complex (6.3%) and the ST-677 complex (6.3%) (Table
1). Despite this, however, the ST-45 lineage was found to be
highly diverse and comprised of various sequence types with
similar prevalence. Likewise, the distribution of isolates among
sequence types of other lineages was fairly even throughout the
data set, while the overall population structure of C. jejuni
isolates in wild birds appeared to be highly diverse (Table 1).
This feature was somewhat unique to the wild bird data set,
since it has been observed that C. jejuni isolates from other
sources often belong to dominant sequence types within clonal
lineages, including the ST-45 complex. However, this observa-
tion may support the suggestion that the relatively high body
temperature of birds could provide an optimal growth envi-
ronment for campylobacters (19), and therefore, isolates may
be genetically more diverse (25). Adding to this observation
was the presence of a large group of genetically distinct isolates
with novel sequence types that were unassigned to clonal com-
plexes (35.9%) (Fig. 2), which contributed to 52% of all such
isolates of the entire data set. This observation may also be a
reflection of the high diversity of C. jejuni isolates in wild birds
but warrants further investigation. The ST-45 complex (5, 9,
25) and unassigned sequence types (25) were found to be
highly prevalent in poultry sources in previous studies, and the
results of this study suggest that this may also be true in wild
birds. The coincidence of such findings in poultry and wild
birds could suggest that similar mechanisms may exist in the
gastrointestinal tract of avian animals to facilitate the niche
adaptation of similar C. jejuni strains.
Given the association of the ST-45 complex between poultry
and human infections (9, 25), the high prevalence of the ST-45
complex observed in this data set strongly suggests that wild
birds may play a part in the role of disease transmission and
should also be regarded as a potential reservoir for C. jejuni
infections. Additionally, seven other complex strains with
lower prevalence in the wild bird data set have also caused
human gastroenteritis in the past. However, these observations
are in contrast to previous studies that demonstrated a signif-
icantly limited degree of overlap between C. jejuni genotypes
found in certain wild bird species and those found in human
clinical disease using pulsed-field gel electrophoresis (2, 3).
Possible explanations for inconsistencies may include geo-
graphical differences, typing methods used, and the difference
in bird taxa investigated between studies, where it has been
demonstrated that C. jejuni strains could differ according to
species and/or specific feeding habits and ecology (3).
The C. jejuni genotypes found in environmental water
closely resembled those found in the wild bird data set, where
the ST-45 complex and unassigned sequence types were pre-
dominant. Unlike in birds, however, these C. jejuni isolates
appeared to be less diverse, as the ST-45 complex only in-
cluded two sequence types and presented fewer unassigned
sequence types. This may indicate that, while environmental
water is contaminated with C. jejuni isolates from wild animals,
only a proportion of genotypes were more adapted to survive
and persist in the environment. Further, the low diversity of
sequence types observed in water may also be, in part, a re-
flection of the fact that Campylobacter spp. do not replicate
outside hosts (21), hence the lack of means by which genetic
variation could be generated following contamination.
There are implications for human infection from the finding
that six of the eight clonal complexes isolated from environ-
mental water in this area matched those that have caused
disease in humans and that geographical areas such as the
location studied are used for recreational activities such as
water sports, camping, and picnicking. This suggestion, how-
5136 KWAN ET AL.APPL. ENVIRON. MICROBIOL.
ever, also conflicts with two previous studies that were con-
ducted in New Zealand, which have concluded that genotypes
found in environmental water did not overlap with those that
cause human disease, although different characterization
methods were used in these studies (7, 11).
The wild mammal isolates included in this study consisted
mainly of rabbit and a number of badger samples, where the
ST-45 complex was predominant, although the ST-21 clonal
complex and unassigned sequence types were also highly prev-
alent in rabbits. Interestingly, rabbit was the only source where
there was comparable prevalence between a genotype that was
found to be dominant in cattle (the ST-21 complex) and ge-
notypes that were found to be dominant in wild birds and water
(ST-45 and unassigned sequence types). In addition, similar to
that found in water samples, the ST-45 complex from these
sources was also less diverse and had fewer unassigned se-
quence types than that found in wild birds. The most prevalent
genotypes in the wild mammal data set were also the most
relevant to human infections, in particular, clonal complexes
ST-45, ST-21, ST-61, ST-42 (9), and ST-677 (20), which col-
lectively represented 73.7% of rabbit isolates.
The associations of clonal complexes between isolation
sources were compared, and distinct as well as overlapping
genotypes were identified among different sources within the
study area. For example, the ST-61 complex was significantly
associated with cattle, while the ST-21 complex was also found
in significantly higher numbers compared to that found in
other sources except rabbits. A set of common genotypic char-
acteristics among wildlife and environmental isolates that was
distinct from cattle isolates was also apparent, which was the
high prevalence of clonal complexes ST-45, ST-677, and ST-
952 and unassigned sequence types.
The finding that ST-45 complex isolates were significantly
associated with sources from wildlife and the environment is in
contrast with the suggestion from previous studies that it is
predominantly a poultry-adapted strain. This may indicate that
the ST-45 complex was also widespread in the natural environ-
ment, or it may be a reflection of contamination of the envi-
ronment from animal sources. The latter, however, may be a
more plausible explanation; wild birds could be a likely source
of environmental contamination considering the hypothesis
that, as previously discussed, the ST-45 complex may be an
avian-adapted C. jejuni strain (5, 25), although additional stud-
ies would be necessary to form robust conclusions. Nonethe-
less, the high prevalence of the ST-45 complex observed in
both wild birds and in environmental water suggests that this
strain is capable of withstanding marked differences between
the high temperatures of birds and the ambient temperatures
of environmental waters.
In addition, despite the relatively lower prevalence, the ST-
677 and ST-952 complex strains were predominantly isolated
from wildlife and environmental water sources, and only one
isolate belonging to the ST-677 complex was found in cattle.
While the ST-952 complex was a newly identified lineage in
this study, in a previous study, infection with the ST-677 com-
plex was found to be associated with drinking unchlorinated
water or water from natural sources (20). An increase in the
sample size may reveal a higher number of these strains from
water and wildlife sources.
A considerable number of uncommon and genetically dis-
tinct sequence types, many of which were newly identified in
this study, have been observed from wildlife and environmental
sources. This observation is in-line with a previous study, where
infections with novel and unassigned sequence types were as-
sociated with swimming in natural bodies of water (20). There-
fore, it is suspected that these isolates may in fact represent a
group of closely related isolates for which the genetic links are
yet to be identified due to the small numbers recovered to date.
Further studies on a larger scale may be able to identify the
relationship between these strains and genotypes that are com-
monly found in the environment, such as the ST-952 complex,
or as new emerging C. jejuni clonal groups associated with the
The finding of low numbers of the ST-21 complex from
wildlife and environmental sources was somewhat surprising,
since it has been considered to be a stable cluster of C. jejuni
isolates that is ubiquitous and has adapted to a wide range of
hosts and environments (5, 9, 25). However, due to the limited
number of environmental isolates that has been investigated
using MLST to date, the question of whether the ST-21 com-
plex is indeed less widespread in the natural environment de-
serves further research.
The associations between clonal complexes and sources
were confirmed by performing a gene flow analysis based on
3,309-bp-long concatenated nucleotide sequences from all
seven housekeeping loci (Table 3). Considering that an FST
value of 0.932 was found between the two Campylobacter spe-
cies (C. jejuni and C. coli) in a previous study (8), the FST
values of 0.199 and 0.175 found within C. jejuni isolates are
noteworthy and are indicative of a limited gene flow between
C. jejuni isolate populations in cattle with those in wild birds
and the environment, respectively. This is in agreement with
the segregated distribution of clonal complexes observed be-
tween these sources, suggesting that transmission from cattle
to humans is perhaps through nonenvironmental routes, that
is, through the food chain.
Our primary objective was to investigate the potential role of
wildlife and the environment as a source of C. jejuni infection
in cattle by comparing the distribution of genotypes from di-
verse sources. These observations have demonstrated that C.
jejuni isolate populations in wildlife and environmental sam-
ples were notably distinct from those in cattle within this geo-
graphical area. The finding that certain clonal complexes were
predominantly shared among isolates from wildlife and envi-
ronmental sources is indicative of common or interconnecting
horizontal transmission pathways of C. jejuni isolates within
these sources. However, the dissemination of genotypes be-
tween cattle and the environment and wildlife appeared to be
considerably restricted, and we suggest that C. jejuni isolates
from the environment have limited significance as reservoirs
for infection of the cattle population currently under study.
From the public health viewpoint, the finding that the ma-
jority of the C. jejuni isolate genotypes identified in this study
have been associated with human infections emphasizes the
need to recognize cattle and their associated environment as
important potential reservoirs for human disease, in particular
for clonal complex strains ST-21, ST-45, and ST-61. Further-
more, the distinct distribution of these genotypes among sam-
ples from different sources observed in this study would imply
that their transmission to humans is perhaps via independent
VOL. 74, 2008 MOLECULAR EPIDEMIOLOGY OF C. JEJUNI5137
routes. This situation would have implications in the design of Download full-text
Investigations that have addressed the importance of the
potential transmission of C. jejuni isolates between cattle and
environmental sources have been scarce. Based on the system-
atic sampling of isolates in this study, a population snapshot
that mirrored the dynamics of C. jejuni isolates within a dairy
farmland microcosm was explored with MLST, a characteriza-
tion tool proven to be valuable for discerning the molecular
epidemiology of C. jejuni. This has offered further insight into
C. jejuni epidemiology, where both widely distributed geno-
types and potentially host-associated genotypes have been
identified. The added knowledge and understanding of such
relationships would further our ability to recognize the sources
of campylobacteriosis. To corroborate these findings, and
given the high degree of significance of the C. jejuni genotypes
identified in this study with respect to human disease, further
studies involving a larger number of isolates and focusing on
wildlife and environmental samples should be an essential re-
The funding of this work was supported by the Health Protection
Agency North West, United Kingdom.
This study made use of the Campylobacter jejuni multilocus sequence
typing website (http://pubmlst.org/campylobacter/) developed by Keith
Jolley and Man-Suen Chan and sited at the University of Oxford (15).
The development of this site has been funded by the Wellcome Trust.
1. Bolton, F. J., S. B. Surman, K. Martin, D. R. A. Wareing, and T. J. Hum-
phrey. 1999. Presence of Campylobacter and Salmonella in sand from bathing
beaches. Epidemiol. Infect. 122:7–13.
2. Broman, T., H. Palmgren, S. Bergstro ¨m, M. Sellin, J. Waldenstro ¨m, M. L.
Danielsson-Tham, and B. Olsen. 2002. Campylobacter jejuni in black-headed
gulls (Larus ridibundus): prevalence, genotypes, and influence on C. jejuni
epidemiology. J. Clin. Microbiol. 40:4594–4602.
3. Broman, T., J. Waldenstro ¨m, D. Dahlgren, I. Carlsson, I. Eliasson, and B.
Olsen. 2004. Diversities and similarities in PFGE profiles of Campylobacter
jejuni isolated from migrating birds and humans. J. Appl. Microbiol. 96:834–
4. Brown, P. E., O. F. Christensen, H. E. Clough, P. J. Diggle, C. A. Hart, S.
Hazel, R. Kemp, A. J. H. Leatherbarrow, A. Moore, J. Sutherst, J. Turner,
N. J. Williams, E. J. Wright, and N. P. French. 2004. Frequency and spatial
distribution of environmental Campylobacter spp. Appl. Environ. Microbiol.
5. Colles, F. M., K. Jones, R. M. Harding, and M. C. J. Maiden. 2003. Genetic
diversity of Campylobacter jejuni isolates from farm animals and the farm
environment. Appl. Environ. Microbiol. 69:7409–7413.
6. Corry, J. E. L., and H. I. Atabay. 2001. Poultry as a source of Campylobacter
and related organisms. Symp. Ser. Soc. Appl. Microbiol. 30:96S–114S.
7. Devane, M. L., C. Nicol, A. Ball, J. D. Klena, P. Scholes, J. A. Hudson, M. G.
Baker, B. J. Gilpin, N. Garrett, and M. G. Savill. 2005. The occurrence of
Campylobacter subtypes in environmental reservoirs and potential transmis-
sion routes. J. Appl. Microbiol. 98:980–990.
8. Dingle, K. E., F. M. Colles, D. Falush, and M. C. J. Maiden. 2005. Sequence
typing and comparison of population biology of Campylobacter coli and
Campylobacter jejuni. J. Clin. Microbiol. 43:340–347.
9. Dingle, K. E., F. M. Colles, R. Ure, J. A. Wagenaar, B. Duim, F. J. Bolton,
A. J. Fox, D. R. A. Wareing, and M. C. J. Maiden. 2002. Molecular charac-
terization of Campylobacter jejuni clones: a basis for epidemiologic investi-
gation. Emerg. Infect. Dis. 8:949–955.
10. Dingle, K. E., F. M. Colles, D. R. A. Wareing, R. Ure, A. J. Fox, F. E. Bolton,
H. J. Bootsma, R. J. L. Willems, R. Urwin, and M. C. J. Maiden. 2001.
Multilocus sequence typing system for Campylobacter jejuni. J. Clin. Micro-
11. Eyles, R. F., H. J. L. Brooks, C. R. Townsend, G. A. Burtenshaw, N. C. K.
Heng, R. W. Jack, and P. Weinstein. 2006. Comparison of Campylobacter
jejuni PFGE and Penner subtypes in human infections and in water samples
from the Taieri River catchment of New Zealand. J. Appl. Microbiol. 101:
12. French, N., M. Barrigas, P. Brown, P. Ribiero, N. Williams, H. Leatherbar-
row, R. Birtles, E. Bolton, P. Fearnhead, and A. Fox. 2005. Spatial epidemi-
ology and natural population structure of Campylobacter jejuni colonizing a
farmland ecosystem. Environ. Microbiol. 7:1116–1126.
13. Furtado, C., G. K. Adak, J. M. Stuart, P. G. Wall, H. S. Evans, and D. P.
Casemore. 1998. Outbreaks of waterborne infectious intestinal disease in
England and Wales, 1992–5. Epidemiol. Infect. 121:109–119.
14. Gonzalez, I., K. A. Grant, P. T. Richardson, S. F. Park, and M. D. Collins.
1997. Specific identification of the enteropathogens Campylobacter jejuni and
Campylobacter coli by using a PCR test based on the ceuE gene encoding a
putative virulence determinant. J. Clin. Microbiol. 35:759–763.
15. Hudson, J. A., C. Nicol, J. Wright, R. Whyte, and S. K. Hasell. 1999. Seasonal
variation of Campylobacter types from human cases, veterinary cases, raw
chicken, milk and water. J. Appl. Microbiol. 87:115–124.
16. Jolley, K. A., M. S. Chan, and M. C. J. Maiden. 2004. mlstdbNet—distrib-
uted multi-locus sequence typing (MLST) databases. BMC Bioinformatics
17. Jones, K. 2001. Campylobacters in water, sewage and the environment.
J. Appl. Microbiol. 90:S68–S79.
18. Jones, K., M. Betaieb, and D. R. Telford. 1990. Thermophilic campylobacters
in surface waters around Lancaster, UK: negative correlation with Campy-
lobacter infections in the community. J. Appl. Bacteriol. 69:758–764.
19. Kapperud, G., and O. Rosef. 1983. Avian wildlife reservoir of Campylobacter
fetus subsp. jejuni, Yersinia spp., and Salmonella spp. in Norway. Appl. En-
viron. Microbiol. 45:375–380.
20. Ka ¨renlampi, R., H. Rautelin, D. Scho ¨nberg-Norio, L. Paulin, and M. L.
Ha ¨nninen. 2007. Longitudinal study of Finnish Campylobacter jejuni and C.
coli isolates from humans, using multilocus sequence typing, including com-
parison with epidemiological data and isolates from poultry and cattle. Appl.
Environ. Microbiol. 73:148–155.
21. Ketley, J. M. 1997. Pathogenesis of enteric infection by Campylobacter.
22. Kwan, P. S. L., A. Birtles, F. J. Bolton, N. P. French, S. E. Robinson, L. S.
Newbold, M. Upton, and A. J. Fox. 2008. Longitudinal study of the molecular
epidemiology of Campylobacter jejuni in cattle on dairy farms. Appl. Environ.
23. Leatherbarrow, A. J. H., C. A. Hart, R. Kemp, N. J. Williams, A. Ridley, M.
Sharma, P. J. Diggle, E. J. Wright, J. Sutherst, and N. P. French. 2004.
Genotypic and antibiotic susceptibility characteristics of a Campylobacter coli
population isolated from dairy farmland in the United Kingdom. Appl.
Environ. Microbiol. 70:822–830.
24. Linton, D., R. J. Owen, and J. Stanley. 1996. Rapid identification by PCR of
the genus Campylobacter and of five Campylobacter species enteropathogenic
for man and animals. Res. Microbiol. 147:707–718.
25. Manning, G., C. G. Dowson, M. C. Bagnall, I. H. Ahmed, M. West, and D. G.
Newell. 2003. Multilocus sequence typing for comparison of veterinary and
human isolates of Campylobacter jejuni. Appl. Environ. Microbiol. 69:6370–
26. Millson, M., M. Bokhout, J. Carlson, L. Spielberg, R. Aldis, A. Borczyk, and
H. Lior. 1991. An outbreak of Campylobacter jejuni gastroenteritis linked
to meltwater contamination of a municipal well. Can. J. Public Health 82:
27. Petersen, L., E. M. Nielsen, J. Engberg, S. L. W. On, and H. H. Dietz. 2001.
Comparison of genotypes and serotypes of Campylobacter jejuni isolated
from Danish wild mammals and birds and from broiler flocks and humans.
Appl. Environ. Microbiol. 67:3115–3121.
28. Roberts, J. A., P. Cumberland, P. N. Sockett, J. Wheeler, L. C. Rodrigues, D.
Sethi, and P. J. Roderick. 2003. The study of infectious intestinal disease in
England: socio-economic impact. Epidemiol. Infect. 130:1–11.
29. Rozas, J., J. C. Sa ´nchez-DelBarrio, X. Messeguer, and R. Rozas. 2003.
DnaSP, DNA polymorphism analyses by the coalescent and other methods.
30. Smith, A., M. Reacher, W. Smerdon, G. K. Adak, G. Nichols, and R. M.
Chalmers. 2006. Outbreaks of waterborne infectious intestinal disease in
England and Wales, 1992–2003. Epidemiol. Infect. 134:1141–1149.
31. Sockett, P. 1993. Social and economic aspects of food-borne disease. Food
32. Stanley, K., and K. Jones. 2003. Cattle and sheep farms as reservoirs of
Campylobacter. J. Appl. Microbiol. 94:104S–113S.
33. Tauxe, R. V. 1992. Epidemiology of Campylobacter jejuni infections in the
United States and other industrial nations, p. 9–12. In I. Nachamkin, M. J.
Blaser, and L. S. Tompkins (ed.), Campylobacter jejuni: current and future
trends. American Society for Microbiology, Washington, DC.
34. Waldenstro ¨m, J., T. Broman, I. Carlsson, D. Hasselquist, R. P. Achterberg,
J. A. Wagenaar, and B. Olsen. 2002. Prevalence of Campylobacter jejuni,
Campylobacter lari, and Campylobacter coli in different ecological guilds and
taxa of migrating birds. Appl. Environ. Microbiol. 68:5911–5917.
5138 KWAN ET AL.APPL. ENVIRON. MICROBIOL.