Epidemiol. Infect. (2016), 144, 3326–3334.

© Cambridge University Press 2016

doi:10.1017/S0950268816001539

A role for flies (Diptera) in the transmission of Campylobacter

to broilers?

A. ROYDEN 1 *, A. WEDLEY 1 , J. Y. MERGA 1 , S. RUSHTON 2 , B. HALD 3 ,

T. HUMPHREY 1 A N D N. J. WILLIAMS 1
1
Department of Epidemiology and Population Health, Institute of Infection and Global Health, University of
Liverpool, Liverpool, UK
2
School of Biology, University of Newcastle, Newcastle-upon-Tyne, UK
3
National Food Institute, Technical University of Denmark, Søborg, Denmark

Received 25 February 2016; Final revision 10 June 2016; Accepted 14 June 2016;
first published online 15 August 2016

SUMMARY
Campylobacter is the leading cause of bacterial diarrhoeal disease worldwide, with raw and
undercooked poultry meat and products the primary source of infection. Colonization of broiler
chicken flocks with Campylobacter has proved difficult to prevent, even with high levels of
biosecurity. Dipteran flies are proven carriers of Campylobacter and their ingress into broiler
houses may contribute to its transmission to broiler chickens. However, this has not been
investigated in the UK. Campylobacter was cultured from 2195 flies collected from four UK
broiler farms. Of flies cultured individually, 0·22% [2/902, 95% confidence interval (CI) 0–0·53]
were positive by culture for Campylobacter spp. Additionally, 1293 flies were grouped by family
and cultured in 127 batches: 4/127 (3·15%, 95% CI 0·11-6·19) from three broiler farms were
positive for Campylobacter. Multilocus sequence typing of isolates demonstrated that the flies
were carrying broiler-associated sequence types, responsible for human enteric illness. Malaise
traps were used to survey the dipteran species diversity on study farms and also revealed up to
612 flies present around broiler-house ventilation inlets over a 2-h period. Therefore, despite the
low prevalence of Campylobacter cultured from flies, the risk of transmission by this route may
be high, particularly during summer when fly populations are greatest.

Key words: Broiler chicken, Campylobacter, Diptera, flies, foodborne zoonoses.

I N T RO D U C T I O N the UK only one in ten cases are reported to national

Campylobacter is the leading cause of human bacterial surveillance schemes [2]. Aside from possible compli-
diarrhoeal disease worldwide. In the European Union cations of acute infection, campylobacteriosis can
(EU) alone, 214 779 cases were reported in 2013 [1]. lead to a number of chronic sequelae [3]. These sign-
However, due to under-reporting, this figure underes- ificant public health implications are coupled to a ser-
timates the true impact of the disease. For example, in ious socioeconomic burden, with campylobacteriosis
costing the EU 0·35 million disability-adjusted life-
years and €2·4 billion per annum [4].
* Author for correspondence: Miss A. Royden, Department of The majority (>90%) of cases are caused by
Epidemiology and Population Health, Institute of Infection and C. jejuni, with the second most frequently isolated spe-
Global Health, University of Liverpool, Leahurst Campus,
Cheshire, UK, CH64 7TE.
cies, C. coli, accounting for a further 5–10% of cases
(Email: [email protected]) [5]. Both species colonize the intestinal mucosa of
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which
permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
 Campylobacter transmission by flies 3327

food-producing animals, including poultry [3]. Raw M AT E R IA L S AN D M E T H O D S

chicken meat can have very high (>107 cells per car- Data collection 1
cass) contamination levels [6] and may account for
up to 70% of cases of campylobacteriosis [7]. An Collection and identification of flies
EU-wide survey conducted on Campylobacter in Flies were caught on four broiler farms within a 2·5
broiler chickens at slaughter, found 75·3% of broiler mile radius in North Wales, UK, from July to
batches colonized with Campylobacter in the UK, August 2011. The four broiler farms consisted of
Europe’s largest producer of broiler meat [8]. The pre- three or four broiler houses stocking about 26 000–
vention of broiler flock colonization has become a 30 000 broilers per house per flock cycle. All four
European food safety priority. farms operated an all-in/all-out system, with thinning
However, Campylobacter colonization of broiler at around day 35 and a 36- to 42-day crop cycle.
flocks is difficult to prevent even if good biosecurity The houses operated a tunnel ventilation system.
is maintained, particularly during the summer All four broiler farms were part of food assurance
months. Campylobacteriosis in temperate regions, schemes requiring adherence to a number of strict
such as the UK, has distinct seasonality, with the biosecurity protocols, including, but not limited to,
peak incidence of human cases occurring in summer restricted and monitored access, an anteroom at
[9]. This correlates with the peak in broiler the entrance to the broiler house containing a phys-
Campylobacter colonization, which occurs in late ical barrier delineating a biosecure area, widespread
spring and summer [10]. This peak in flock coloniza- use of disinfectant footbaths, farm- and broiler-
tion remains unexplained and hypotheses include sea- house-specific clothing and footwear and rigorous
sonal changes in climatic factors, which concurrently policies preventing the introduction and spread of
result in peak fly (Diptera) populations during sum- disease. The four farms were in close proximity to
mer months [10–12]. livestock, including ruminants and horses. Live flies
It has been established previously that flies are car- were collected for culture during at least one whole
riers of Campylobacter spp. and are proven vectors of crop cycle on each farm, by repeat visits once or
other enterobacterial pathogens [13–15]. Flies act as twice weekly. At each sampling event, flies were in-
mechanical vectors, transmitting microorganisms as dividually collected in sterile re-sealable plastic
fomites, and as biological vectors, multiplying patho- bags outside broiler houses (<10 m periphery) for
gens within the gut [13, 16]. Campylobacter spp. are 2 h. An open bag was placed into a plastic jar and
common in the farm environment and livestock faeces the jar placed over a fly resting on a surface to cap-
represent a major environmental source [17, 18]. ture it. The bag containing the fly was removed from
Studies conducted within the poultry farm en- the jar and transported to the laboratory for
vironment have demonstrated the carriage of processing.
Campylobacter spp. by flies [13, 19, 20]. In addition, Captured flies were anaesthetized through place-
flies have been shown to transmit C. jejuni to chickens ment at −18 °C for 10 min and identified taxonomic-
under laboratory conditions [21], and molecular typ- ally to family level under a microscope. Dipteran
ing in a Danish field study strongly implicated flies families were categorized as (i) filth flies, (ii) livestock,
in the transmission of a strain of C. jejuni from dung or carrion associated flies and (iii) other flies.
sheep to broilers [22]. Further, Danish studies strongly Filth flies commonly consume and breed in excreta
suggest that the ingress of flies into broiler houses con- and carrion and include those of the families
tributes to the transmission and colonization of broiler Calliphoridae, Fanniidae and Muscidae [14]. Where
chickens with Campylobacter spp. [23, 24]. possible, the species of Calliphoridae, Fanniidae or
While data are available from several countries to Muscidae were determined, as these have previously
support the hypothesis that flies are implicated in the been most highly implicated in the carriage of
epidemiology of Campylobacter colonization of broil- Campylobacter spp. [12, 14, 22]. Category (ii) contains
er flocks, field studies have not previously taken place families comprising a large proportion of species de-
in the UK and there are no data on the carriage of pendent upon animal excrement or organic matter
Campylobacter spp. by flies on and around UK broiler for part of their lifecycle, i.e. mating, oviposition or
farms. Therefore, this study aimed to investigate the feeding. The families contained in category (iii) are
role of flies in the transmission of Campylobacter generally not associated with livestock, dung or
spp. to broiler chickens in the UK. carrion.
3328 A. Royden and others

Bacterial isolation and identification June to August 2012. As described, flies were collected
Following identification, each fly was individually and on farm and transported to the laboratory. Captured
manually macerated for 10 s, enriched in 10 ml mod- flies were anaesthetized with CO2 gas and identified
ified Exeter selective enrichment broth [1100 ml nutri- taxonomically to family level. While anaesthetized,
ent broth, 11 ml lysed defibrinated horse blood, flies were grouped into batches of ∼10 flies of similar
Campylobacter enrichment supplement SV59 (Mast dipteran families using sterile forceps.
Group Ltd, UK) and Campylobacter growth supple-
ment SV61 (Mast Group Ltd)] and incubated under Bacterial isolation and identification
microaerobic conditions at 42 °C. After 24 h and 48 h
incubation, 5 µl of enrichment broth was streaked Following grouping of flies into batches, microbio-
onto Campylobacter-selective blood-free agar [modified logical culture and PCR of Campylobacter spp. was
charcoal-cefoperazone-deoxycholate agar (mCCDA)] carried out as in data collection 1.
supplemented with cefoperazone (32 mg/l) and ampho-
tericin B (10 mg/l). These plates were incubated micro- Broiler chicken flock sampling and testing for
aerobically at 42 °C for 48 h and up to four colonies Campylobacter
per plate were selected based on morphology and
streaked onto Columbia agar containing 5% (v/v) One broiler chicken flock on each of the four farms
defibrinated horse blood. Plates were incubated for was tested daily for Campylobacter from June to
48 h under both microaerobic conditions at 42 °C and August 2012. To sample the broiler flock, one pair
aerobic conditions at 37 °C, to distinguish morpho- of disposable fabric boot socks (overshoes) was
logically similar Campylobacter and Arcobacter species. worn over rubber boots as the sampler walked 100 m
Four Campylobacter isolates were stored at −80 °C in inside the broiler house. Boot socks were pre-
Microbank™ vials (Pro-Lab, UK). All media were moistened with sterile physiological saline to allow
obtained from Lab M Ltd (UK) and all blood from maximum uptake of Campylobacter from the litter.
Southern Group Labs (UK). Each pair of boot socks was then placed into an indi-
Genus- and species-specific polymerase chain reac- vidual sterile re-sealable plastic bag for transport. In
tion (PCR) assays were performed on the four selected the laboratory, 200 ml sterile physiological saline
suspected Campylobacter colonies. To extract DNA, was added to each pair of boot socks and agitated
∼10 µl of bacterial cells were suspended in 300 µl chelex by hand to release attached matter. Boot socks and sa-
solution [20% w/v in 100 ml Trizma HCl (Sigma- line were left to settle for 10 min, then 1 ml clear
Aldrich, UK)] and heated at 95 °C for 10 min, centri- upper fluid was transferred to a sterile 1·7 ml
fuged at 16200 g for 3 min and 50 µl of the clear super- Eppendorf tube and centrifuged at 16200 g for 7
natant added to 450 µl sterile distilled water. A min. The supernatant was discarded and DNA
confirmatory multiplex 16S rDNA PCR assay for extracted from the pellet using the Promega
Campylobacter and Arcobacter spp. [25, 26] was fol- Genomic DNA Wizard Extraction kit (Promega,
lowed by a multiplex C. jejuni and C. coli specific UK). Presence of Campylobacter was confirmed by
PCR assay [27]. Each PCR reaction consisted of 21 µl performing PCR as in data collection 1. To confirm
of 1·1 × (2·5 mM MgCl2) ReddyMix PCR Master Mix flock positivity following a positive PCR result, 30
(Thermo Scientific, UK), 1 µl of bovine serum albumin swabs of excreted chicken faeces were collected from
(Sigma-Aldrich), 0·25 µl of each primer and 2 µl tem- the broiler flock. Faecal swabs were streaked directly
plate DNA. DNA amplification was performed for onto mCCDA and incubated microaerobically at
30 cycles using an annealing temperature of 50 °C. 42 °C for 48 h. If <28 swabs were Campylobacter-
PCR amplicons were visualized after electrophoresis positive by culture, a further 30 faecal swabs were col-
on a 2% agarose gel (Alpha Laboratories, UK). lected every other day until at least 28 positive swabs
were obtained. Additionally, the flock continued to be
boot-sock sampled daily until Campylobacter had
Data collection 2 been isolated from at least 28 faecal swabs. Finally,
ten Campylobacter isolates were randomly selected
Collection and identification of flies from these final 28–30 cultures to represent ten indi-
A second sampling phase was conducted on the same vidual faecal samples. Isolates were stored at −80 °C
four broiler farms sampled in data collection 1 from in Microbank™ vials (Pro-Lab).
 Campylobacter transmission by flies 3329

Multilocus sequence typing (MLST) analysis Data collection 2

To assess genetic diversity, MLST profiles were obtained Culture and MLST of Campylobacter spp.
for C. jejuni and C. coli isolates. From the four isolates
Flies were sampled on each broiler farm 5–6 times, to-
stored from each Campylobacter-positive batch of flies talling 21 sampling events on all four farms. A mean
and ten isolates from each Campylobacter-positive of 62 flies were caught during each sampling event.
flock, two isolates were randomly selected for MLST
A total of 1293 flies, representing 29 families, were
from each positive batch of flies or broiler flock. These collected and cultured in 127 batches with a mean of
were subjected to the Campylobacter-specific MLST ten flies per batch (range 1–17) (Supplementary
scheme as described by Miller et al. [28], with alternative Table S2). Batch size varied due to grouping of similar
PCR and sequencing primers used for the tkt locus families of Diptera to aid taxonomical identification
where necessary [29]. Sequence data were analysed of dipteran species carrying Campylobacter spp.
using ChromasPro (Technelysium Pty Ltd, Australia) C. jejuni was isolated from four batches of flies from
and allele numbers and sequence types (STs) assigned the broiler farms (4/127, 3·15%, 95% CI 0·11-6·19)
using the Campylobacter MLST website (developed by (Table 1). One positive fly batch (farm D) contained
Keith Jolley and sited at the University of Oxford; only species of the family Calliphoridae, while the
http://pubmlst.org/campylobacter/) [30].
other three positive batches contained a variety of spe-
cies from dipteran families associated with livestock,
Malaise trap survey of flies (Diptera) dung or carrion, including the filth-fly families,
Malaise traps were used for community analysis to Calliphoridae, Fanniidae and Muscidae.
survey the species and numbers of Diptera gaining ac- Three C. jejuni sequence types (STs) were identified
cess to the broiler houses through the ventilation inlets from the four C. jejuni-positive batches of flies caught
and to analyse the diversity of dipteran species present on three broiler farms, with ST1701 [clonal complex
on the study farms. A Malaise trap is a tent-like struc- (CC) 45] identified on two farms (B and D). During
ture, where flies and other insects fly into the tent wall the study period, a total of six flock cycles occurred
and are funnelled up to the highest point and into a across the four broiler farms, all of which were
collection vessel. Two Malaise traps were placed per- Campylobacter-positive before thinning. From these
pendicular to broiler-house walls, at a right angle to six flocks, MLST profiles were obtained for two iso-
the insect flight line, at each sampling event for 2 h. lates from each of three Campylobacter-positive
Seventy per cent ethanol was used in the collection flocks. These six allelic profiles revealed two STs,
vessel as a killing agent and preservative prior to with both isolates from each flock identified as the
identification. same ST. The broiler flock at farm A tested positive
for C. coli [ST828 (CC828)] at 21 days of age.
However, no Campylobacter-positive flies were caught
R E SU LTS on this farm. On farm D, the broilers were found posi-
Data collection 1 tive for C. jejuni on day 13 [ST257 (CC257)] and two
batches of flies caught on day 26 were positive for two
Culture of Campylobacter spp. different C. jejuni STs of the same CC (ST25 and
Flies were sampled on each broiler farm 6–9 times, to- ST1701 of CC45). On farms B and C, two flock cycles
talling 29 sampling events on all four farms. A mean were covered during the study. On farm B, the first
of 31 flies were caught during each sampling event. flock during the study period was positive for C. jejuni
A total of 902 flies, representing 28 families, were col- (unknown ST) at 21 days; 6 days before C. jejuni
lected and cultured individually (Supplementary [ST1701 (CC45)] was cultured from a batch of flies.
Table S1). Two flies were found to be carrying The second flock was C. jejuni-positive [ST257
Campylobacter spp.; a Calliphora vomitoria or ‘blue- (CC257)] at 15 days; however, no Campylobacter-
bottle’ was positive for C. lari and a fly of the positive flies were caught during this flock cycle.
Heleomyzidae, a dipteran family associated with live- During the first flock cycle of the study period on
stock, dung or carrion, was found to be carrying C. farm C, the flock was positive for C. jejuni (unknown
coli. The prevalence of Campylobacter spp. from ST), but no positive flies were caught. During the se-
flies was calculated to be 0·22% [2/902, 95% con- cond flock cycle, C. jejuni [ST137 (CC45)] was iden-
fidence interval (CI) 0-0·53]. tified in a batch of flies at 2 days and the flock was
Table 1. Batches of flies (Diptera) from four broiler farms in the UK testing positive for Campylobacter spp. between June and August 2012 (data collection 2)

3330
and STs obtained through MLST of Campylobacter isolates from flies and broiler flocks

Flock age when flies Campylobacter sp. Flock age when broilers Campylobacter sp. isolated

A. Royden and others

Campylobacter-positive isolated from flies Campylobacter-positive from flock [ST (CC)] [flock
Farm Sampling date (days) [ST (CC)] Flies present in positive batch (days) age at isolation (days)]

A August 2012 Flies Campylobacter- Flies Campylobacter- Flies Campylobacter-negative 21 C. coli [ST828 (CC828)]
negative negative (21)
B June 2012 27 C. jejuni 7 filth flies (1 Fanniidae, 1 Muscina 21 C. jejuni
(flock cycle 1) [ST1701 (CC45)] stabulans, 3 Phaonia sp., 2 Polietes (ST unknown) (29)
lardarius); 3 livestock, dung and carrion
associated flies (1 Anthomyiidae, 1
Psychodidae, 1 Scatophagidae); 2 other
flies (1 Dolichopodidae, 1 unidentified
sp.)
August 2012 Flies Campylobacter- Flies Campylobacter- Flies Campylobacter-negative 15 C. jejuni [ST257 (CC257)]
(flock cycle 2) negative negative (21)
C June 2012 Flies Campylobacter- Flies Campylobacter- Flies Campylobacter-negative 35 C. jejuni
(flock cycle 1) negative negative (ST unknown) (37)
July 2012 2 C. jejuni 3 filth flies (1 Fanniidae, 1 Muscidae, 21 Unidentified
(flock cycle 2) [ST137 (CC45)] 1 Phaonia sp.); 7 livestock, dung Campylobacter sp.
and carrion associated flies (3 (ST unknown) (21)
Anthomyiidae, 3 Scatophagidae,
1 Sciaridae)
D July 2012 26 C. jejuni 10 filth flies (8 Calliphora vicina, 13 C. jejuni
[ST25 (CC45)] 1 Calliphora vomitoria, 1 Phormia [ST257 (CC257)] (15)
terraenovae)
26 C. jejuni 9 filth flies (1 Calliphora vicina, 5 Fannia 13 C. jejuni
[ST1701 (CC45)] canicularis, 1 Phaonia sp., 2 Stomoxys [ST257 (CC257)] (15)
calcitrans); 1 livestock, dung and carrion
associated flies
(1 Scatophagidae)

ST, Sequence type; CC, clonal complex; MLST, multilocus sequence typing.
 Campylobacter transmission by flies 3331

positive for an unidentified Campylobacter spp. at 21 Additionally, a Swedish study discovered 1·0% (3/
days. 291) of flies carrying Campylobacter spp. in the prox-
imity of 31 broiler farms [31].
Malaise trap survey of flies (Diptera) In data collection 2, isolates from Campylobacter-
positive broiler flocks and batches of flies were typed
Two Malaise traps were set up during 20 sampling using MLST. No direct correlation of STs was
events, collecting a total of 1771 insects (class: found between isolates from flies and broilers.
Insecta), including 1644 flies, from 28 dipteran fam- However, the STs of three flock isolates could not
ilies (Table 2, Supplementary Table S3). The range be determined as amplicons were not produced with
of Diptera caught in one sampling event was 0–612. any MLST loci primers. One C. jejuni ST [ST257
(CC257)] and one C. coli ST [ST828 (CC828)] were
identified in three flocks from the broiler farms. The
D I S C US S IO N STs identified in the broiler flocks belong to chicken-
This is the first study to investigate the prevalence of associated CCs [32]. Three C. jejuni STs (ST25,
Campylobacter spp. carried by flies on UK commer- ST137 and ST1701) belonging to CC45 were identified
cial broiler farms and demonstrates that flies are car- in four batches of flies from three broiler farms. CC45
riers of Campylobacter in this environment. The is associated with human enteric illness and food
prevalence of Campylobacter spp. was calculated as animal populations, particularly chicken flocks [32].
0·22% (2/902, 95% CI 0–0·53) from flies cultured indi- While flies were only found positive for
vidually in data collection 1 and 3·15% (4/127, 95% CI Campylobacter before the broiler flock on one occa-
0·11–6·19) from batches of flies cultured in data collec- sion (Table 1), this indicates that flies may transmit
tion 2. Following data collection 1, it was decided to Campylobacter to humans either directly or indirectly
group flies into batches for logistical purposes. via broiler chickens.
However, this prevented determining prevalence at Transmission of Campylobacter spp. from flies to
the individual fly level. In data collection 1, a fly of broilers may be successful if a source exists within a
the Heleomyzidae was found to be carrying C. coli short distance of the broilers and there is a consistently
and a Calliphora vomitoria was positive for C. lari. high turnover rate of new flies acquiring the pathogen
This indicates that positive batches may be the result [33, 34]. This is viable in the UK due to the general
of a single fly carrying Campylobacter, suggesting proximity of pastured livestock to broiler farms. The
that the prevalence of Campylobacter from flies is four broiler farms in this study were all in close prox-
lower than the estimated 3·15% batch prevalence. imity to livestock, including ruminants and horses.
There was one minor difference in the microbiological Given the association of CC45 with food animal
methodology for the two data collections; in data populations, the origin of the C. jejuni isolated from
collection 1, flies were anaesthetized by being placed flies in this study may have been the surrounding live-
at −18 °C for 10 min; however, in data collection 2, stock. Previous studies have provided evidence that
flies were anaesthetized with CO2 gas. This difference flies are carriers of Campylobacter over short dis-
in anaesthetic technique is not anticipated to have tances; Hald et al. [22] discovered the same macro-
affected the rate of recovery of Campylobacter spp. restriction PFGE fingerprint in 27/28 C. jejuni isolates
from flies. Ultimately, carriage of Campylobacter by from broilers, flies and sheep on one farm and Stern
flies is at low frequencies in the broiler-farm et al. [35] found flaA type 15 in both flies and chicken
environment. intestinal samples from one broiler farm. A
Few studies have caught flies in numbers compar- Norwegian study found a clustering of broiler flocks
able to this study. However, prevalence estimates up to 4 km apart positive for Campylobacter spp. in
compare favourably to European studies with similar summer, which indicates the presence of factors acting
sample sizes. Hald et al. [12] caught 2186 flies around on a narrow scale, such as flies [11].
five Danish broiler farms and found 31 (1·1%) to The Malaise traps captured a high diversity of flies,
be positive for Campylobacter spp. However, this totalling 28 families of Diptera. The Malaise traps do
estimate may have been influenced by a single not confirm which fly families and species are entering
broiler farm, which also had a pig farm on the same broiler houses and in what frequencies. However, up
site, as very few positive flies were found (<1·0%) to 612 flies were caught adjacent to the broiler-house
on the four farms without additional livestock. ventilation inlets in just 2 h, presenting multiple
3332 A. Royden and others

Table 2. Number of insects (class: Insecta) and flies (Diptera) caught in Malaise traps on four broiler farms in the
UK between June and August 2012 (data collection 2)

Broiler farm A Broiler farm B Broiler farm C Broiler farm D Broiler farm total

Filth flies 1 (0·8) 8 (0·7) 6 (1·7) 10 (4·9) 25 (1·4)

Livestock, dung or carrion associated 79 (65·3) 989 (90·7) 212 (59·9) 123 (59·7) 1403 (79·2)
flies
Other flies 34 (28·1) 61 (5·6) 75 (21·2) 46 (22·3) 216 (12·2)
Non-Diptera 7 (5·8) 32 (2·9) 61 (17·2) 27 (13·1) 127 (7·2)
Diptera (total) 114 (94·2) 1058 (97·1) 293 (82·8) 179 (86·9) 1644 (92·8)
Overall total 121 (100) 1090 (100) 354 (100) 206 (100) 1771 (100)

Values given are n (%).

opportunities for flies to enter the house. This sup- inlets. MLST demonstrated that flies commonly
ports the findings of other studies [12, 22, 31], which carry broiler-associated STs, responsible for human
also conclude that despite the low prevalence of enteric illness. Flies threaten the microbiological
Campylobacter spp. from flies, risk of transmission is safety of the food chain, with an ability to breach
potentially high as flies are present in large numbers existing biosecurity measures and colonize broiler
around broiler-house ventilation inlets. It has been flocks. Future studies should aim to quantify the risk
estimated that the average influx of insects into a of colonization that flies present to UK broiler farms
broiler house per broiler rotation is 30 728 insects and implement strategies to reduce this risk. In
(range 2233–180 300) [12]. Twenty-one per cent of Denmark, fly screens have been shown to significantly
these were flies, of which 20·3% were filth flies and (P = 0·0002) reduce the prevalence of Campylobacter-
the remaining majority composed of families that positive flocks at slaughter [23] and the use of a similar
are attracted to livestock, dung or carrion. A similar case-control study in the UK would be beneficial in
distribution was seen in this study; where in total, assessing the risk that flies present and the effective-
86·9% of flies caught in Malaise traps on broiler ness of available control measures.
farms were associated with livestock, dung or carrion;
however, only 1·5% were filth flies. Increased ventila-
tion airflow during warmer weather also increases
the likelihood of insects being carried passively into S U P P L EM E N TA RY MAT E R I A L
houses [10]. Additionally, if the Campylobacter preva- For supplementary material accompanying this paper
lence from flies is compared to other vector-borne dis- visit http://dx.doi.org/10.1017/S0950268816001539.
eases, similarly low prevalences are observed, which
indicates that the prevalence seen in this study may
be sufficient for flies to be a regular source of col-
onization for the broiler flock. For example, 0·114% AC K N OWL E D G E M E NT S
(n = 70 937) of flies were carrying Shigella spp. in We thank Elli Wright, Ruth Ryvar and Cathy Glover
Southwestern USA where flies were heavily implicated for their technical assistance and staff from the study
in the epidemiology of local cases [36]. farms for their cooperation. The research leading to
In summary, Campylobacter spp. were isolated these results has received funding from the European
from flies caught on broiler farms and this study pro- Union’s Seventh Framework Programme (FP7/2007–
vides the first estimate of the prevalence of 2013) (grant no. 244547); and The Wellcome Trust
Campylobacter spp. from flies on UK broiler farms. Clinical Veterinary Research Training Programme
This study demonstrates that flies may play a role in (A.R., grant no. 082161).
the transmission of Campylobacter to broilers.
Despite the low prevalences of Campylobacter spp.
detected from flies on broiler farms (0·22% and
3·15%), risk of transmission is high as flies are present D E C L A R AT I O N O F I N T E R E S T
in large numbers around broiler-house ventilation None.
 Campylobacter transmission by flies 3333

R E FE RE NC E S 15. Wright EP. The isolation of Campylobacter jejuni from

flies. Journal of Hygiene 1983; 91: 223–226.
1. European Food Safety Authority. The European Union 16. Ekdahl K, Normann B, Andersson Y. Could flies explain
summary report on trends and sources of zoonoses, zoo- the elusive epidemiology of campylobacteriosis? BMC
notic agents and food-borne outbreaks in 2013. EFSA Infectious Diseases 2005; 5: 11.
Journal 2015; 13: 3991–4155. 17. Stanley KN, et al. Seasonal variation of thermophilic
2. Tam CC, et al. Longitudinal study of infectious intes- campylobacters in lambs at slaughter. Journal of
tinal disease in the UK (IID2 study): incidence in the Applied Microbiology 1998; 84: 1111–1116.
community and presenting to general practice. Gut 18. Stanley K, Jones K. Cattle and sheep farms as reservoirs
2012; 61: 69–77. of Campylobacter. Journal of Applied Microbiology
3. Skirrow MB. Diseases due to Campylobacter, 2003; 94 (Suppl. 1): 104–113.
Helicobacter and related bacteria. Journal of 19. Berndtson E, Danielsson-Tham M, Engvall A. Campylo-
Comparative Pathology 1994; 111: 113–149. bacter incidence on a chicken farm and the spread of
4. European Food Safety Authority Panel on Biological Campylobacter during the slaughter process.
Hazards. Scientific opinion on Campylobacter in broiler International Journal of Food Microbiology 1996; 32:
meat production: control options and performance 35–47.
objectives and/or targets at different stages of the food 20. Gregory E, et al. Epidemiological study of Campylobac-
chain. EFSA Journal 2011; 9: 2105–2245. ter spp. in broilers: source, time of colonization, and
5. Gillespie IA, et al. A case-case comparison of prevalence. Avian Diseases 1997; 41: 890–898.
Campylobacter coli and Campylobacter jejuni infection: 21. Shane SM, Montrose MS, Harrington KS. Transmission
a tool for generating hypotheses. Emerging Infectious of Campylobacter jejuni by the housefly (Musca domes-
Diseases 2002; 8: 937–942. tica). Avian Diseases 1985; 29: 384–391.
6. Jørgensen F, et al. Prevalence and numbers of 22. Hald B, et al. Flies and Campylobacter infection of
Salmonella and Campylobacter spp. on raw, whole broiler flocks. Emerging Infectious Diseases 2004; 10:
chickens in relation to sampling methods. International 1490–1492.
Journal of Food Microbiology 2002; 76: 151–164. 23. Hald B, Sommer HM, Skovgård H. Use of fly screens to
7. Boysen L, et al. Source attribution of human campylo- reduce Campylobacter spp. introduction in broiler
bacteriosis in Denmark. Epidemiology & Infection houses. Emerging Infectious Diseases 2007; 13: 1951–
2014; 142: 1599–1608. 1953.
8. European Food Safety Authority. Analysis of the base- 24. Bahrndorff S, et al. Foodborne disease prevention and
line survey on the prevalence of Campylobacter in broil- broiler chickens with reduced Campylobacter infection.
er batches and of Campylobacter and Salmonella on Emerging Infectious Diseases 2013; 19: 425–430.
broiler carcasses in the EU, 2008 – Part A: 25. Linton D, Owen RJ, Stanley J. Rapid identification by
Campylobacter and Salmonella prevalence estimates. PCR of the genus Campylobacter and of five
EFSA Journal 2010; 8: 1503–1602. Campylobacter species enteropathogenic for man and
9. Nylen G, et al. The seasonal distribution of animals. Research in Microbiology 1996; 147: 707–718.
Campylobacter infection in nine European countries 26. Gonzalez I, et al. Development of a combined PCR-
and New Zealand. Epidemiology and Infection 2002; culture technique for the rapid detection of Arcobacter
128: 383–390. spp. in chicken meat. Letters in Applied Microbiology
10. Humphrey T, O’Brien S, Madsen M. Campylobacters as 2000; 30: 207–212.
zoonotic pathogens: a food production perspective 27. Klena JD, et al. Differentiation of Campylobacter coli,
[Review]. International Journal of Food Microbiology Campylobacter jejuni, Campylobacter lari, and
2007; 117: 237–257. Campylobacter upsaliensis by a multiplex PCR devel-
11. Jonsson ME, et al. Effect of climate and farm environ- oped from the nucleotide sequence of the lipid A gene
ment on Campylobacter spp. colonization in Norwegian lpxA. Journal of Clinical Microbiology 2004; 42: 5549–
broiler flocks. Preventative Veterinary Medicine 2012; 5557.
107: 95–104. 28. Miller WG, et al. Extended multilocus sequence typing
12. Hald B, et al. Influxed insects as vectors for Campylo- system for Campylobacter coli, C. lari, C. upsaliensis,
bacter jejuni and Campylobacter coli in Danish broiler and C. helveticus. Journal of Clinical Microbiology
houses. Poultry Science 2008; 87: 1428–1434. 2005; 43: 2315–2329.
13. Rosef O, Kapperud G. House flies (Musca domestica) as 29. Dingle KE, et al. Multilocus sequence typing system for
possible vectors of Campylobacter fetus subsp. jejuni. Campylobacter jejuni. Journal of Clinical Microbiology
Applied and Environmental Microbiology 1983; 45: 2001; 39: 14–23.
381–383. 30. Jolley K, Maiden M. BIGSdb: Scalable analysis of bac-
14. Szalanski AL, et al. Detection of Campylobacter and terial genome variation at the population level. BMC
Escherichia coli O157:H7 from filth flies by polymerase Bioinformatics 2010; 11: 595.
chain reaction. Medical and Veterinary Entomology 31. Hansson I, et al. Correlations between Campylobacter
2004; 18: 241–246. spp. prevalence in the environment and broiler
3334 A. Royden and others

flocks. Journal of Applied Microbiology 2007; 103: 34. Kjærsgaard A, et al. Plasticity in behavioural responses
640–649. and resistance to temperature stress in Musca domestica.
32. Sheppard SK, et al. Campylobacter genotypes from food Animal Behaviour 2015; 99: 123–130.
animals, environmental sources and clinical disease 35. Stern NJ, et al. Flagellin A gene restriction fragment length
in Scotland 2005/6. International Journal of Food polymorphism patterns of Campylobacter spp. isolates from
Microbiology 2009; 134: 96–103. broiler production sources. Avian Diseases 1997; 41: 899–905.
33. Skovgård H, Kristensen K, Hald B. Retention of Cam- 36. Richards CS, et al. Studies on rates of recovery of
pylobacter (Campylobacterales: Campylobacteraceae) Shigella from domestic flies and from humans in
in the house fly (Diptera: Muscidae). Journal of Medical Southwestern United States. American Journal of
Entomology 2011; 48: 1202–1209. Tropical Medicine and Hygiene 1961; 10: 44–48.