Campylobacters in water, sewage and the environment

Authors


K. Jones Department of Biological Sciences, IENS, Lancaster University, Lancaster, LA1 4YQ, UK (e-mail: k.jones@lancaster.ac.uk).

1. Summary, 68S

2. Introduction, 68S

3. Campylobacters in water, 68S

3.1 Fresh water, 69S

3.1.1 Streams, 69S

3.1.2 Rivers, 69S

3.1.3 Canals, 70S

3.1.4 Ponds and ornamental lakes, 70S

3.1.5 Groundwater, 70S

3.1.6 Reservoirs, 70S

3.1.7 Drinking water, 70S

3.2 Marine, 71S

3.2.1 Coastal waters/bathing waters and sediment, 71S

3.2.2 campylobacters for coastal bathing waters, 71S

3.2.3 Estuaries, 71S

4. Campylobacters in sewage, 72S

4.1 Sewage treatment, 72S

4.2 Effects of sewage effluent on receiving surface waters, 72S

4.3 Sewage sludge, 72S

4.3.1 Sewage sludge put to land, 73S

5. Correlation with faecal indicators, 73S

6. Miscellaneous aspects of survival, 73S

6.1 Protection in biofilms, 74S

6.2 Viable but non-culturable Campylobacter, 74S

6.3 Survival of an infective dose, 74S

7. Environmental loading and seasonality, 74S

7.1 The farm as an environmental reservoir of infection, 74S

7.2 Shedding of campylobacters in faeces of livestock, 74S

7.3 Campylobacters in farm waste, 75S

7.4 Wild animals, wild birds and pets, 75S

7.5 Are campylobacter strains found in the environment important?, 75S

8. Conclusions, 76S

9. Acknowledgements, 76S

10. References, 76S

1. SUMMARY

Thermophilic campylobacters are widespread in the environment, where they are a sign of recent contamination with animal and avian faeces, agricultural run-off and sewage effluent. Although intestinal carriage of campylobacters is ubiquitous in livestock, domestic animals, wild animals, wild birds and poultry, contamination of the environment with the bacteria in faeces is intermittent and varies seasonally, depending on factors such as stress and changes in diet. Wild birds, and not sewage effluent, are the source of campylobacters in some coastal waters. The density of Campylobacter spp. in sewage effluent depends on the source of the sewage and the type of treatment. There is a qualitative, but not a quantitative, correlation between campylobacters and faecal indicators in environmental samples. The marked seasonal pattern of campylobacters in temperate, aquatic environments is a result of variations in Campylobacter die-off rates at different times of the year.

2. INTRODUCTION

The presence of campylobacters in environmental samples can be taken as a sign of recent faecal contamination, because not only are campylobacters unable to multiply outside warm-blooded host animals, they also survive for a shorter time than the usual indicators, faecal coliforms and faecal streptococci (Bolton et al. 1987; Korhonen and Martikainen 1991; Wallace et al. 1993; Koenraad et al. 1997; Obiri-Danso and Jones 1999a).

3. CAMPYLOBACTERS IN WATER

The incidence of enteritis caused by thermophilic campylobacters has the hallmarks of a waterborne disease, in that there is a major spring peak followed by a trough in the summer and a small peak in autumn. However, this is not backed up by quantitative studies of surface waters, where numbers are higher in the winter than in the summer (Fig. 1).

Figure 1.

 Negative correlation between hours of sunshine (▵) and campylobacter numbers in Morecambe Bay (black columns) and the Lune estuary (white columns)

3.1. Fresh water

3.1.1. Streams.

The presence of thermophilic campylobacters in streams varies with location, season and agricultural practice. Studies of streams in north-west England have shown that campylobacters are absent from streams running through upland moors but present in the same streams running through lowland, grazed pasture (Jones et al. 1990a; Jones and Hobbs 1996). The composition of the Campylobacter population is dependent on the path of the stream (Obiri-Danso and Jones 1999a). Streams running through pasture contain mainly C. jejuni with some C. coli, shed by grazing cattle (Stanley et al. 1998c) and sheep (Stanley et al. 1998d; Jones et al. 1999b), whereas those draining duck ponds contained a mixture of C. jejuni, C. lari, C. coli and urease-positive thermophilic campylobacters (UPTCs), which are typical of avian sources. A further study showed that campylobacters occurred intermittently in streams, with their density correlating with upstream agricultural locations (Table 1), such as farmyards, small-holdings and a slaughterhouse, and agricultural events, such as emptying of slurry tanks and the spraying of farm slurry onto land (Jones and Hobbs 1996; Stanley et al. 1998b). These observations confirm the generally held view that agricultural slurries, manure and municipal sewage following land disposal, or unintentional leakage from sewers, farm stockpiles or slurry holding pits, are the primary sources of contamination of streams with microorganisms of public health concern (Goss and Barry 1995; Watson 1985).

Table 1.   Examples of event-led and location-led agricultural contamination of streams with CampylobacterThumbnail image of

3.1.2. Rivers.

Thermophilic campylobacters are ubiquitous in rivers, especially those exposed to agricultural run-off and effluent from water treatment plants (Bolton et al. 1987; Jones et al. 1990a; Stelzer et al. 1991; Jones and Hobbs 1996; Popowski et al. 1997; Obiri-Danso and Jones 1999a). A detailed study of the river Conder (north-west England) showed that the upper, cleaner reaches of the river were free of Campylobacter but, as the water flowed through grazed farmland, it became progressively more contaminated with campylobacters, particularly C. jejuni (Fig. 2.) (Jones et al. 1990a; Jones and Hobbs 1996). A 2-year survey of two inland recreational waters on the river Lune in the same geographical area (Obiri-Danso and Jones 1999a) confirmed the results reported by Bolton et al. (1987) and Jones et al. (1990a) which showed that campylobacters were present all year round, but with lower numbers in the summer corresponding to elevated UV levels and higher temperatures. Sources of Campylobacter contamination for the Lune included: sewage effluent from a small waste water treatment works (secondary treatment by trickle filtration), streams containing agricultural run-off, sheep and cattle grazing meadows bordering the river and an indigenous population of ducks. C. jejuni was the predominant species isolated; however, the presence of C. lari, C. coli and UPTCs is indicative of a range of sources. There was little diurnal variation (numbers did not decline during the day), which reflects the diverse and continuous nature of the inputs. The river sediments showed the same distribution of species but the numbers remained constant throughout the year, showing no seasonality. A study of rivers and lakes in the Warsaw region of Poland showed that 70% of water samples were positive for thermotolerant Campylobacter with C. jejuni making up 65% of the isolates, C. coli 22% and C. lari 13%. Municipal sewage was stated to be the main source, with minor inputs from the droppings of wild animals (Popowski et al. 1997).

Figure 2.

 Increase in the density of campylobacters in the river Conder from freshwater source to estuary

3.1.3. Canals.

The Lancaster canal in north-west England contains low numbers of campylobacters. Their density peaks in the winter and contamination originates from streams, run-off from pasture and ducks (Jones et al. 1990a).

3.1.4. Ponds and ornamental lakes.

Ponds and ornamental lakes, such as the Serpentine in Hyde Park, London, can be of poor microbiological quality and are rich sources of campylobacters, which are shed by the large populations of ducks and wildfowl. For example, water samples taken from a small lake at Lancaster University contained between 23 and 430 campylobacters per 100 ml of water, and 30% of faecal samples, shed by a variety of ducks, waterfowl and peacocks, were positive for Campylobacter. Faecal samples from mallard ducks (Anas platyrhinchos) contained between 1·6 × 106 and 2·7 × 107 campylobacters per g dry weight, and of the 82 isolates, 90% were C. jejuni, 7% C. coli and 3% UPTCs.

3.1.5. Groundwater.

Groundwater is widely considered to be microbiologically clean. It rises to the surface at boreholes, wells, springs or seeps and is frequently used for drinking water for livestock on farms. Contaminated groundwater sources have been implicated in the introduction of Campylobacter into poultry flocks (Pearson et al. 1993) and broiler chickens (VandeGiessen et al. 1996). However, the first culturable evidence that Campylobacter can occur in groundwater was provided by Stanley et al. (1998a). Hydrological evidence suggested that the source of contamination was a dairy farm situated within the hydrological catchment of the groundwater, and this was confirmed when identical strains of C. jejuni were isolated from groundwater and the dairy herd. The ability to culture Campylobacter from these aquifers suggests that the vertical movement of this organism may significantly threaten the quality of subsurface aquifers and supports the conclusions of Rollins (1991) and Pearson et al. (1993), that contaminated groundwater may be an overlooked source of campylobacters in animals reared for food. The environmental conditions found within subsurface aquifers, i.e. low redox potentials, the absence of molecular oxygen with increasing depth, all year round low temperatures and protection from the effects of UV and desiccation, favour the survival of Campylobacter. Since similar conditions occur in large aquifers delivering water to cities, it is possible that these could be a vehicle for the transmission of campylobacters shed by sheep and birds, as was recently the case with Cryptosporidium in north-west England (Anon 1999).

3.1.6. Reservoirs.

There are reports from Holland that Campylobacter originating from birds was detected in pristine reservoirs, but only in winter months (Medema and Schets 1994). In the most pristine reservoirs, Campylobacter numbers were similar to those for Escherichia coli. Similar observations were made for impounding reservoirs in north-west England (Jones et al. 1990a; Jones et al. 1999a), where campylobacters were absent in late spring and early summer but present in relatively large numbers in winter. Phenotypic profiling was used to trace the origin of the bacteria from three suspect sources, sheep and Canada geese grazing the edge of the reservoir and a colony of black-headed gulls roosting on an island. The sheep and Canada geese excreted only C. jejuni whereas the black-headed gulls excreted C. jejuni, C. lari, C. coli and UPTCs, which matched the profile of species found in the water column.

3.1.7. Drinking water.

The numbers of sporadic infections and outbreaks of Campylobacter enteritis caused by drinking water have recently been reviewed (Koenraad et al. 1997; Thomas et al. 1999). World wide, Koenraad et al. (1997) reported nine outbreaks between 1978 and 1991, each affecting between 13 and 3000 people; Thomas et al. (1999) reported a further seven outbreaks between 1990 and 1997, affecting between 8 and 633 people. They predominantly arose from consumption of untreated or contaminated water. This appears to be a particular problem in Scandinavian countries (Aho et al. 1989), where there is, apparently, a divine right to drink from streams which appear to be clean (Gunnarsson, pers. comm.). Koenraad et al. (1997) state that exposure to contaminated water causes between 1·2 and 170 Campylobacter infections per 100 000 people. In the UK, data on waterborne disease and water are collated by the PHLS Communicable Disease Surveillance Centre and published twice yearly in the Communicable Disease Report (Anon 2000). Their results show that outbreaks of Campylobacter enteritis derived from drinking water are confined to private water supplies, although the association between disease symptoms and Campylobacter is frequently ‘probable’ as the bacterium is seldom isolated from the water source. Such private water supplies are predominantly found in small rural systems which are more likely to be contaminated with animal waste (Wyn-Jones 2000). Mégraud and Serceau (1990) surveyed the water distribution network of a large urban community in France and were unable to find thermophilic campylobacters. Of the 14 non-thermophilic strains isolated, 13 came from untreated water and all were sensitive to chlorination. Hanninen et al. (1998) demonstrated that dairy cows shed more campylobacters in the summer, when the cattle were drinking lake water, than in the winter, when they were provided with chlorinated mains water. However, a similar indoor/outdoor pattern was not detected in a study of dairy and beef cattle in north-west England (Stanley et al. 1998c).

Campylobacters have been shown to be widespread in surface waters in Kumasi, Ghana. Interestingly, samples of drinking water taken from a water hole in the early morning contained campylobacters, whereas samples taken in the late afternoon did not, presumably due to the biocidal effects of high temperatures and elevated levels of UV radiation (Adjei, pers. comm.).

3.2. Marine

3.2.1. Coastal waters/bathing waters and sediment.

Studies carried out in Morecambe Bay (north-west England) show that campylobacters are widespread in coastal water (Jones et al. 1990a; Obiri-Danso and Jones 1999b), where they show a distinct seasonal variation, i.e. they are more numerous in the winter and either absent or present in only low numbers in the summer (Fig. 1). In the bathing season (May to September) campylobacters exhibited diurnal variation, i.e. seawater samples were positive for campylobacters in the early morning and negative in the late afternoon (Obiri-Danso et al. 1999). Both the seasonal distribution and the diurnal variation showed statistically significant correlations with the levels of UV radiation and changes in temperature (Jones et al. 1990a, 1991; Obiri-Danso and Jones 1999b; Obiri-Danso et al. 1999, 2001). The predominant Campylobacter spp. found in Morecambe Bay seawater were C. lari and UPTC (Obiri-Danso and Jones 1999b). A recent study, using natural populations, showed that C. lari and UPTCs survived for longer in seawater than either C. jejuni or C. coli (Obiri-Danso et al. 2001).

Thermophilic campylobacters were also found in the surface layer of the intertidal sediments in Morecambe Bay, where they displayed similar seasonal patterns and species composition (C. lari and UPTC) to those in the overlying sea water (Obiri-Danso and Jones 2000). However, Bolton et al. (1999) compared Campylobacter contamination of sand on EEC standard and non-standard bathing beaches, and showed that campylobacters could be isolated from 50% of sand samples from non-EEC standard beaches and 40% from EEC standard beaches, with C. jejuni and C. coli more prevalent on non-EEC standard beaches and C. lari and UPTC on EEC standard beaches.

3.2.2. Sources of campylobacters for coastal bathing waters.

It had been assumed that the main source of campylobacters for bathing waters was sewage effluent, even though the bacteria in effluent survived for only a short time at ambient temperatures in seawater (Jones et al. 1990a, [34]b). The introduction of a waste water treatment works (WWTW) at Morecambe in north-west England provided an opportunity to test whether this assumption was valid. Prior to 1997, Morecambe’s sewage had been macerated and discharged via a short sea outfall into the ebb tide. After 1997 the sewage was subject to secondary treatment and discharged via a long sea outfall several miles south of the original outfall. Seawater samples were taken twice monthly for a year both before and after the treatment plant came on-line, and tested for thermophilic campylobacters and faecal indicators. The introduction of sewage treatment resulted in a reduction in the density of faecal indicators in the bathing waters but not in the density of campylobacters (Obiri-Danso and Jones 1999a). It has also been shown that the species composition of campylobacters in the seawater and sediments of Morecambe Bay is made up of 20–40% C. lari and 60–80% UPTC, whereas in sewage effluent it consists of 83% C. jejuni and 17% C. coli (Jones et al. 1999b; Obiri-Danso and Jones 1999c, 2000). As Morecambe Bay supports large flocks of wild birds such as oystercatchers, knots and various types of gull, which excrete large numbers of Campylobacter with the same profile of species (C. lari and UPTC) found in the seawater and sediments (Jones et al. 1998; Fitzgerald et al. 1998), it is probable that birds and not sewage are the source of campylobacters for these coastal bathing waters.

3.2.3. Estuaries.

Estuaries have a similar seasonal pattern of campylobacters as coastal waters (Jones et al. 1990a). A study of the Lune estuary in north-west England confirmed that Campylobacter numbers were higher in the winter than in the summer, that they increased towards the mouth of the estuary and that densities were higher at high water than low water (Zainuldin and Jones 1996). There were a variety of sources of campylobacters, namely, the fresh water section of the river, various streams and contaminated overflows, Lancaster’s WWTW, sheep and cattle grazing the salt marshes, a large gullery on an urban waste tip and the incoming seawater. A full range of Campylobacter species were isolated, with C. jejuni, C. coli, C. lari and UPTCs present in roughly equal proportions, depending on when and where in the estuary samples were taken. Samples taken at low water, furthest from the estuary mouth and close to the gullery, had the same species composition (C. jejuni, C. coli, C. lari and UPTC) as the gulls, which move onto the water and forage on the intertidal sediments of the river at low water. At the mouth of the estuary the species composition depended on when samples were taken. Incoming seawater contained C. lari and UPTCs, whereas samples taken 4 h after high water (coinciding with a plume of sewage effluent from Lancaster’s WWTW reaching the mouth of the estuary) contained C. jejuni. Cockles and mussels at the mouth of the estuary contained the same species of Campylobacter as found in seawater and oystercatchers, namely C. lari and UPTCs (Jones et al. 1999b).

4. CAMPYLOBACTERS IN SEWAGE

Results from several continents have shown that campylobacters are ubiquitous in sewage and that humans and animal waste from abattoirs and animal treatment plants are the major sources (Arimi et al. 1988; Höller 1988; Stelzer et al. 1988, 1991; Jones et al. 1990a, [34]b; Betaieb and Jones 1990; Stampi et al. 1992; Koenraad et al. 1994; Lauria-Filgueiras and Hofer 1998; Stampi et al. 1999). Analysis of individual sewers in Lancaster, UK, showed that Campylobacter numbers were related to the incidence of campylobacteriosis in subdivisions of the city and to the presence of animal effluent from abattoirs and poultry processing ([34]Jones et al. 1990b). Campylobacters in sewage showed an identical seasonal pattern to that of the human population (Fig. 3), namely, a large peak in late May and June and a minor one in September or October (Jones et al. 1990a, [34]b; Jones and Telford 1991). In Italy, Stampi et al. (1992) also found higher counts in May, June and July. In 1989, 73% of the year’s input of campylobacters into Lancaster’s sewage occurred in May and June compared with only 29% of Lancaster’s cases of campylobacteriosis. This led to the hypothesis that annual peaks in human infections are related to increased amounts of Campylobacter in the environment, which are in turn determined by changes in the numbers of Campylobacter within livestock, poultry and wild animals (Jones and Telford 1991).

Figure 3.

 Seasonal variation in Campylobacter infections in Lancaster (□) and the numbers of campylobacters in sewage effluent (black columns) and undigested sewage sludge (white columns) in the same year

4.1. Sewage treatment

Treatment by primary settlement in Lancaster’s WWTW produced effluent containing between 262 and 79 000 campylobacters per 100 ml of effluent, depending on the time of the year. Arimi et al. (1988) showed that primary settlement resulted in reductions of 78% in the number of campylobacters.

Reports in the literature for the effects of secondary treatment of sewage by trickle filtration, activated sludge and oxidation ponds on the numbers of Campylobacter, show reductions of between 88% and 99% (Stelzer et al. 1988; Stampi et al. 1992; Lauria-Filgueiras and Hofer 1998). In the UK, reductions of between 86% and 99% were obtained for a large WWTW at Salford (Greater Manchester), of between 79% and 99% for a rotary biofilm treatment system at Middleton (Morecambe) and of between 95% and 100% for a small WWTW at Caton. During hot, sunny weather campylobacters did not survive the treatment process in the small trickle filter WWTW at Caton. Höller (1988) found no seasonal differences in Campylobacter numbers in treated effluent in Germany. Tertiary treatment of effluent with chlorine eliminates campylobacters (Betaieb and Jones 1990; Stampi et al. 1992).

Overall, the number of thermophilic campylobacters in the final effluents from WWTWs depends on the type of treatment. Primary settlement reduces the numbers slightly and they still show the same seasonal pattern as the inputs. Secondary treatment reduces the numbers considerably and tertiary treatment completely eliminates them.

4.2. Effects of sewage effluent on receiving surface waters

Sewage effluents from WWTWs discharging into freshwater rivers contribute to the overall loading of campylobacters in the rivers (Bolton et al. 1987; Jones et al. 1990a). However, die-off and dilution significantly reduce the numbers downstream of the discharge point (Obiri-Danso and Jones 1999a).

Research into the seasonality of campylobacters in coastal waters and an estuary receiving untreated sewage discharges, showed that the numbers of culturable campylobacters declined very quickly (Jones et al. 1990a). Natural populations of Campylobacter in sewage effluent became unculturable after only 15 min of direct sunlight ([34]Jones et al. 1990b) and short in situ survival times for natural populations in seawater were demonstrated by Obiri-Danso and Jones (1999b) and Obiri-Danso et al. (1999, 2001). Wallace (1994) showed that the survival of pure cultures of C. jejuni in water depended on the time of year. They survived for only a few minutes during June and July compared with up to 4 h during December and January. Not only were campylobacters more susceptible to high light intensities than other bacteria, but they were unable to repair damage to DNA.

4.3. Sewage sludge

Fresh sewage sludge contains high numbers of campylobacters, which, in the north-west of England, show the same seasonality as human infections (Fig. 3) (Jones et al. 1990c). Sludge digestion normally reduced Campylobacter numbers to zero but occasional positive samples were obtained in the winter. On these occasions it was unclear whether the campylobacters had survived sludge treatment or whether fresh, untreated sludge from another WWTW had been added post-treatment. Other workers have also shown that campylobacters in sludge are eliminated by digestion processes (Stelzer et al. 1991; Stampi et al. 1999). However, Stampi et al. (1999) have also warned that large numbers of Arcobacter survive sludge digestion and thickening and have the potential to contaminate crops.

4.3.1. Sewage sludge put to land.

Although the inability to find measurable numbers of Campylobacter in sludge put to land has been reported (Jones et al. 1990c), in 1998, 8% of swab samples taken from sludge put to land were positive for C. jejuni (unpublished). As sludge production is increasing and the amount put to land in the UK is likely to double over the next few years, the possible contamination of sewage sludge with campylobacters and arcobacters should be re-addressed. With the implementation of The Safe Sludge Matrix (ADAS 1999), the disposal of sewage sludge to land has become tightly regulated, but, as with the disposal of the less well regulated farm slurries, much seems to depend on the integrity of the farmer.

5. CORRELATION WITH FAECAL INDICATORS

It is generally accepted that campylobacters are only found in the presence of faecal coliforms and faecal streptococci. However, there is not always a good correlation between the densities of the Campylobacter population and of the indicators. For example, Carter et al. (1987) isolated campylobacters from a number of natural water sources in central Washington (USA) including ponds, lakes, and small mountain streams, and although campylobacters were recovered throughout the year, especially in autumn and winter months, their densities did not show significant correlation with the faecal indicators. Arvanitidou et al. (1995) examined 86 river and lake samples from eight sampling sites in northern Greece for the presence of pathogens and faecal indicators and they too found that the density of campylobacters could not be predicted by the density of the standard indicator bacteria. At two inland bathing sites on the river Lune (UK) faecal coliforms were poorly correlated with the density of campylobacters (P < 0·212 during 1996 and P < 0·155 during 1997) (Obiri-Danso and Jones 1999a). The density of faecal streptococci showed a reasonable correlation with the density of campylobacters in 1996 (P < 0·051) but not during 1997 (P < 0·513). In Morecambe Bay (UK) there was a poor relationship between indicators and Campylobacter. For example, the installation of a more efficient WWTW and a longer sea outfall reduced the numbers of faecal indicators but had no effect on the numbers of campylobacters (Obiri-Danso and Jones 1999b). In this case it was because the campylobacters and faecal indicators came from different sources, the campylobacters coming from wild birds and the indicators from sewage effluent. Similarly, in the mouth of the Lune estuary, changes in Campylobacter numbers correlated with the incoming tide, while those for indicators were associated with effluent from Lancaster’s WWTW upstream of the sampling site (Zainuldin and Jones 1996). During activated sludge sewage treatment, Stelzer et al. (1988) found a significant reduction (94·5%) in campylobacters but not in indicator bacteria.

On the other hand, Stelzer et al. (1989) showed a high correlation coefficient (P < 0·05) between the numbers of campylobacters and total coliforms in river water polluted by waterfowl and a poultry farm, and Skjerve and Brennhovd (1992), using a multiple logistic regression model to predict the occurrence of C. jejuni/coli related to three groups of index bacteria, showed that faecal coliforms, but not faecal streptococci, were strong predictors for C. jejuni/coli in a water source in southern Norway. In a further study, Brennhovd et al. (1992) used the same logistic regression analysis to demonstrate a highly significant relationship between the prevalence of Campylobacter and the numbers of all three indicator groups (faecal coliforms, faecal streptococci and sulphite-reducing clostridia).

It is not surprising that the populations of indicators and campylobacters do not always correlate given that campylobacters become non-culturable much faster than the indicators and that the origins of the bacteria can differ.

6. MISCELLANEOUS ASPECTS OF SURVIVAL

The growth of Campylobacter in the environment is limited by its thermophilic and micro-aerophilic nature and the organism is susceptible to environmental stress, such as to ambient temperatures and temperature cycling (Blaser et al. 1980; Rollins and Colwell 1986; Höller et al. 1988), atmo- spheric oxygen concentrations (Hoffman et al. 1979), UV radiation (Bolton et al. 1987; Jones et al. 1990a; Jones et al. 1991; Obiri-Danso and Jones 1999a; Obiri-Danso et al. 1999, 2001) and desiccation (Doyle and Roman 1982). In the absence of spores or any robust physiologically based protective mechanism, the main survival strategy for Campylobacter appears to be the production of huge numbers, in the hope that once out in the ‘environment’, enough will survive to infect another host.

6.1. Protection in biofilms

It has been demonstrated that campylobacters can persist, but not grow, in biofilms, where they are protected from biocidal activity (Buswell et al. 1998). However, this may not be widespread as Stanley et al. (1998a) were unable to isolate them from riverine biofilms on rocks, from pipes immersed in rivers or from pipes in groundwater. Similarly, Popowski et al. (1997) were unable to find campylobacters playing a role in biofilm formation on the surfaces of wood, metal and plastic immersed in surface waters.

6.2. Viable but non-culturable Campylobacter

Since the viable but non-culturable (VBNC) form of Campylobacter was first reported (Rollins et al. 1985; Rollins and Colwell 1986) its presence has been demonstrated in a variety of aquatic systems. Yet, its significance for infection of animals and as the cause of disease for humans is still uncertain. For example, in their review, Koenraad et al. (1997) state that ‘The contribution of viable but nonculturable Campylobacter cells in the contamination cycle has been found to be negligible’ whereas, Thomas et al. (1999) sum up their review with the statement: ‘The virulence of VNC forms should be considered to be equivalent to that of the culturable forms, with the added risk that they are not detectable by conventional culturable methods’. Recent research appears to favour the view that VBNC cells of Campylobacter have an important role in the transmission of the disease, at least for certain strains (Cappelier et al. 1999; Lazaro et al. 1999; Tholozan et al. 1999).

6.3. Survival of an infective dose

In the absence of cross contamination, the number of Campylobacter in a carcass or a meat product will depend on the extent of contamination at slaughter together with the ability to survive during processing, transport and storage. Campylobacters survive best in dark, moist and cool conditions (Phebus et al. 1991; Hazeleger et al. 1998; Lee et al. 1998), which are precisely those which are used in the packaging, storage and transport of meat products from processing plant to the refrigerator in the kitchen. These conditions, which are designed to prevent the multiplication of food poisoning bacteria such as Salmonella and E. coli 0157, are those which help to ensure the survival of an infective dose of Campylobacter.

7. ENVIRONMENTAL LOADING AND SEASONALITY

The seasonal pattern of human infections caused by Campylobacter spp. has remained remarkably constant in northern temperate zones over the last 20 years or so. There is a major peak in late spring and early summer and a smaller one in September (Jones and Telford 1991). After several environmental investigations in the north-west of England, the only ‘environment’ which has shown the same seasonal pattern is sewage (Fig. 3) ([34]Jones et al. 1990b, c). However, sewage is unlikely to be a major environmental reservoir for campylobacters because those in sewage effluent survive for only a short time in surface waters (Jones et al. 1990a; Obiri-Danso and Jones 1999b), and they are largely absent from digested sewage sludge put to land (Jones et al. 1990c). As Campylobacter cannot grow in the environment and survival is limited, it has been suggested that changes in the size of the Campylobacter population within livestock and poultry (Jones and Telford 1991) and the vehicles and vectors which transmit Campylobacter from one host to another, play a significant role in the epidemiology of this organism (Stanley et al. 1998a).

7.1. The farm as an environmental reservoir of infection

Quantitative data for the intestinal carriage of campylobacters is available for most farm animals and birds, for example, beef and dairy cattle (Stanley et al. 1998c), sheep (Stanley et al. 1998d), pigs (Weijtens et al. 1997), chickens (Wallace et al. 1997) and turkeys (Wallace et al. 1998). They show that most, if not all, livestock and poultry carry Campylobacter and that the numbers of bacteria can be very large, especially in poultry (Table 3). There are cyclical variations in the size of the Campylobacter populations within the intestines of cattle, sheep and chickens, which may be responsible for different levels of carcass contamination at different times of the year and contribute to the seasonality of infections shown in man. Campylobacters are readily isolated from the rumen of cattle and sheep (Stanley et al. 1998c, d) which suggests that they are continuously ingesting new strains of the bacterium on the farm. Campylobacters from the intestines of farm animals and poultry enter the wider environment as contamination on meat products, slaughterhouse waste discharged into sewers, and as faeces during grazing. Farm visits, especially by children, can result in campylobacteriosis (Dawson et al. 1995), as can consumption of unwashed vegetables contaminated with manure (Park and Sanders 1992).

7.2. Shedding of campylobacters in faeces of livestock

Although most livestock exhibit intestinal carriage of campylobacters, shedding in faeces is intermittent. For example, in sheep at pasture approximately 30% of faecal samples are positive for campylobacters but this rises to 100% after lambing and when sheep are moved to fresh pasture (Jones et al. 1999a). Factors which encourage faecal shedding of campylobacters by cattle, pigs and sheep include: giving birth, weaning, change of pasture, movement outdoors and transport (Weijtens et al. 1997; Wesley et al. 2000; Stanley et al. 1998c, d; Joneset al. 1999b). Factors which reduce shedding include: gestation, a diet of silage or hay and housing indoors. Although lambs and calves are not contaminated with Campylobacter at birth, when they do become infected, they shed the bacteria at higher rates than adult animals (Stanley et al. 1998c; Jones et al. 1999b). Faecal shedding contributes to the environmental load of campylobacters on pasture and in agricultural run-off.

7.3. Campylobacters in farm waste

Cattle produce approximately 50 l of liquid faeces a day, which is deposited on pasture, spread around the farm when cows are brought in for milking and accumulates as slurry or in bedding when cattle are housed indoors in the winter. On a wet day a farm yard is covered in a film of cattle manure which is picked up on farmers’ boots and by farm dogs and cats. It is unsurprising that raw milk frequently contains campylobacters (Hutchinson et al. 1985). A 100-cow dairy herd produces approximately 750 m3 of slurry in a 6-month period (D. Patterson, http://www greenmount.ac.uk). Slurries are collected and stored in large slurry tanks where campylobacters can survive for long periods (Kearney et al. 1993). Slurry from the tanks is disposed of by spreading onto land, where impact varies with the time of year. Stanley et al. (1998b) showed that not only did aerated slurries put to land in summer contain fewer campylobacters than non-aerated slurries put to land in the winter, but the survival times, after the slurry had been applied to land, were much longer in the winter than in the summer, leading to a greater likelihood of contaminated agricultural run-off in winter than in summer. A Code of Good Practice governs the disposal of slurry onto land for the Protection of Water (MAFF 1999), which states that animals should not be grazed on grassland for at least 21 d after slurry has been applied. However, compared to the situation with sewage sludge, the lack of quality control on the disposal of farm slurries put to land is a cause of increasing concern (McCann 1999). Co-disposal of spent, pyrethroid-based sheep dip with farm slurry resulted in the disappearance of campylobacters from the slurry (Semple et al. 2000).

Faecal waste accumulating in bedding used for cattle housed indoors also contains campylobacters which persist for long periods in dung heaps in the farm yard. Such waste contains straw, is lumpy and is disposed by muckspreading onto pasture. Campylobacters are protected in the lumps and survive for longer than they do in slurries, which are added as a relatively thin film.

Faeces from grazing livestock and slurry and muck put to land are attractive to a variety of birds, including gulls, corvines and shellducks, which peck at the material and potentially transfer the contaminating campylobacters over large distances.

Overall, the farm environment acts as a self-perpetuating reservoir of infection with campylobacters cycling between farm animals, wild animal and avian vectors and inanimate vehicles (boots, tractors).

7.4. Wild animals, wild birds and pets

Campylobacters have been found in a variety of wild animals and pets (Fricker and Park 1989; Park et al. 1991; Cabrita et al. 1992) where they serve as environmental reservoirs of infection for water catchments and, in the case of pets, direct contamination of people and their domestic environment. Birds appear to be the main zoonotic reservoir for thermophilic campylobacters, presumably because of their high internal temperature (Kapperud and Rosef 1983; Fricker and Metcalfe 1984; Pacha et al. 1988; Whelan et al. 1988; Ogg et al. 1989; Glunder et al. 1992; Quessy and Messier 1992). Wild birds have been implicated in contaminating drinking water settling tanks with Campylobacter in Greensville, Florida, USA (Sacks et al. 1986) and there is evidence that birds contaminate milk bottles with campylobacters by pecking at the bottle tops, perhaps after moving from cow-pat to milk bottle on the doorstep (Hudson et al. 1990; Southern et al. 1990; Riordan et al. 1993; Palmer and McGuirk 1995). Wild birds also cause aesthetic damage in towns (Mégraud 1987; Casanovas et al. 1995), contaminate shell fish beds (Teunis et al. 1997; Jones et al. 1999c) and are the source of campylobacters in coastal bathing waters (Obiri-Danso and Jones 1999b, 2000; Obiri-Danso et al. 2001). The presence of campylobacters in migratory birds (Pacha et al. 1988) is indicative of the large distances that campylobacters can be transferred.

7.5. Are campylobacter strains found in the environment important?

Although it is now realized that campylobacters are widespread in the environment, the extent of the risk to health is not known. In north-west England we have shown that the same strains of C. jejuni isolated from patients’ stool samples can be found in livestock, poultry, wild birds, farms, sewage and surface waters. However, apart from the consumption of contaminated food, we still do not have a clear route for the transfer of Campylobacter from the environment to the consumer. Concerted programmes of molecular genotyping to match environmental and human strains of Campylobacter, and environmental epidemiology to trace infection routes, are necessary before the health risks can be fully understood.

8. CONCLUSIONS

Thermophilic campylobacters are almost certainly ubiquitous in the intestines of farm animals, poultry, pets, wild animals and wild birds. The size of the Campylobacter population in animals is variable and may be responsible for different levels of carcass contamination at different times of the year.

Shedding of the bacteria in faeces by animals is intermittent and is dependent on diet and stress. The excretion of large numbers of campylobacters in faeces is a significant survival strategy for Campylobacter.

Contamination of the environment is inevitable and varies seasonally with changes in shedding by animals and inputs from human and animal-derived sewage effluent. Sewage treatment substantially reduces the numbers of Campylobacter in sewage effluent.

The presence of campylobacters in environmental samples correlates qualitatively with the presence of faecal indicators, but the origins of Campylobacter and the indicators are not always the same and densities of Campylobacter and indicators frequently do not correlate quantitatively.

Campylobacters cannot multiply outside the animal host and survival in the environment is relatively poor. Survival of Campylobacter is enhanced by moist, cold conditions, such as those found in refrigerated meat and poultry.

It is not known what proportion of the strains of thermophilic Campylobacter found in the environment are pathogenic, nor is there a clear infection route from the environment to the consumer.

9. Acknowledgements

I wish to express my gratitude for the valuable research carried out at Lancaster University by Drs Mailoud Betaieb, Collette FitzGerald, Kwasi Obiri-Danso, Karen Stanley and Joanne Wallace, with funding from the Libyan Government, the Department of Health, the Ghanaian Government, Lancaster City Council and the Ministry of Agriculture, Fisheries and Food, and for the assistance of several M.Sc and final year project students.

Ancillary