In an earlier study ( Stanley et al. 1998a ), it was shown that almost all sheep carry substantial populations of Campylobacter (mainly Camp. jejuni) in their intestines. However, it was noted that at pasture, the number of faecal samples positive for Campylobacter was much lower than the intestinal carriage rate. In order to see whether the shedding of campylobacters was consistent throughout the year and whether the type of grazing had any effect, a year long study was carried out of the excretion rate of campylobacters by three flocks of sheep grazing on three different types of pasture, namely, saltmarsh, fell (upland) and farm. Horizontal transmission between mothers and new-born lambs was also investigated.
The rates at which sheep on different types of pasture shed campylobacters in their faeces were measured over 12 months. Overall, shedding of campylobacters at pasture was between a third and a half of the carriage rate (92%) of the intestines of sheep at slaughter. Shedding was highest during saltmarsh grazing, followed by upland fell and farm grazing. The rate of shedding varied at different times of the year, with the highest rates (100%) coinciding with lambing, weaning, and movement onto new pasture. The lowest rates (0%) occurred when sheep were fed on hay and silage. On the farm, low rates occurred during the whole of gestation, both when the sheep were indoors and outdoors. Campylobacter jejuni was the main species isolated and survived for up to 4 d in sheep faeces. Lambs became colonized by Campylobacter within 1–5 d of being born. Ewes, which were not shedding campylobacters prior to lambing, started to shed after lambing, and ewes which were shedding low numbers of Campylobacter before lambing, increased the numbers of bacteria being shed after lambing.
Materials and methods
Sheep were sampled from three types of grazing, salt marsh (OS grid reference SD 46 59), farmland (SD 50 57) and an upland fell (SD 50 57), all within 5 miles of Lancaster, UK. Sheep on the fell alternated between rough pasture and moorland grazing. The saltmarsh was not grazed in the winter months and on those dates, sheep were sampled on meadows adjacent to the saltmarsh. Sheep on the farmland were brought inside for lambing from January through to March.
Monthly shedding of thermophilic campylobacters in grazing sheep
Shedding rates were determined in grazing sheep between May 1996 and May 1997. On each sample date, swabs were taken early in the morning from 30 fresh sheep droppings at each of the sites to determine the number of animals shedding the bacteria (isolation or carriage rate). Swabs were placed in 5 ml 0·1% buffered peptone water in test tubes and transported to the laboratory. Isolation rates and enumerations of thermophilic campylobacters were carried out using the enrichment procedures, growth conditions and media described in Wallace et al. (1997) and Stanley et al. (1998b) .
Identification of thermophilic campylobacters
Thermophilic campylobacters were presumptively identified by colony formation, catalase and oxidase reactions, and microscopy using Gram stain with carbol fuchsin as the counterstain. They were phenotypically biotyped according to the modified scheme of Bolton et al. (1992) or the MAST ID Camp Identification System (Mast Diagnostics, Bootle, UK).
Colonization of new-born lambs
In the spring of 1995, the rate at which new-born lambs became colonized with Campylobacter was studied. Two groups of lambs (plus mothers) were isolated from the other sheep. Cloacal samples (swabs) from lambs were examined for carriage, and ewe faeces were enumerated for campylobacters.
The average numbers of faecal samples positive for campylobacters in sheep grazing on saltmarsh, lowland farm and upland pasture over a year are shown in Table 1. The lowest shedding rate was on farm pasture and the highest was on saltmarsh pasture. All the sheep at pasture showed an average shedding rate substantially below the carriage rate in the intestines of sheep at slaughter ( Table 1, Fig. 1).
|Pasture||Percentage of samples positive||Number of samples|
|Meadows adjacent to saltmarsh (winter)||33||150|
|Intestinal carriage *||92||360|
In this study, MPN counts of campylobacters in sheep faeces were only done in September 1996. The results ( Table 2) show little difference between the types of pasture, with the numbers on saltmarsh grazing being the highest and on farm pasture the lowest.
|Type of grazing||Number of bacteria g−1 dry wt|
Campylobacter jejuni was the main species (90%) found in grazing sheep, Camp. coli was the next most plentiful (8%) and Camp. lari made up 2%.
Intermittent shedding over the year
Monthly variations in the shedding rate are shown in Figs 1, 2, 3 and 4.
Intestinal carriage of campylobacters.
Campylobacters were found in the intestines of almost all sheep at slaughter ( Fig. 1). The monthly isolation rate was consistently above 90% apart from in February and October.
On upland fell grazing, high shedding rates were found in April (99·5%) and September (100%) ( Fig. 2). The peak in April coincided with lambing and the peak in September, with weaning. In the summer (May to July), shedding was lower on the fells than on saltmarsh or farm pasture. This coincided with the sheep being in much lower densities with much of the flock having been moved onto higher, rougher pasture. In August, the lambs born in April were also moved higher onto the fell. In September, the sheep were brought down from the higher fells onto less rough pasture, and the lambs weaned and sold. Shedding in the autumn and early winter was higher than in the summer and no campylobacters were isolated in January and March. In January, the fells were covered in snow and the sheep were feeding on silage. In March, the sheep had been given an antibacterial vaccine.
On the saltmarsh, the highest shedding rates coincided with lambing and movement onto the saltmarsh from inland pasture (May 1996, April 1997) and weaning of lambs (September). In May 1997, only 17% of adult sheep were shedding compared with 67% for new lambs. Summer shedding rates were higher on the marsh than on either upland or farm pasture. From November 1996 to March 1997, no sheep were grazing the saltmarsh. During this period, sheep (not the same flock as those which had been on the marsh) grazing meadows adjacent to the salt marsh were sampled ( Fig. 2). These sheep were shedding at a higher rate than sheep on the farm or the fells.
Farm grazing (Fig. 3).
Sheep on the farm in the summer showed a steady level of shedding at around 30%. Increased shedding at the time of weaning in September was not detected. Shedding of campylobacters was either low or absent from November through to February. The sheep were taken inside in January (after the January sampling) and fed on hay. In February and March they were fed on silage. By the February sampling date, four ewes had lambed and had been separated from the rest of the flock. Faecal samples of two of these ewes were positive for Campylobacter whereas faecal samples from the rest of the flock were negative. By March, all the ewes had lambed and were in a field next to the farm buildings. By the April sample date they had been put back in their original pasture and the shedding rate reached 100%.
Survival of campylobacters in sheep faeces
Care had been taken to sample fresh faeces early in the morning. However, survival experiments with faecal samples brought back to the laboratory and left outside showed that campylobacters could still be detected in sheep faeces from the fell after 4 d, and in those from the saltmarsh and farm, for 3 d.
The effects of lambing on shedding and the colonization of new-born lambs
In the spring of 1995, the rate at which new-born lambs became colonized with Campylobacter was studied. Two groups of lambs (plus mothers) were isolated from the other sheep and cloacal samples from the lambs were examined for campylobacters. In the first group, colonization was first detected by day 3 and 100% of the lambs were colonized by day 5 ( Fig. 5a). In the second group, colonization was detected on day 1 and reached 100% by day 5 ( Fig. 5b). In the first group, the ewes were not shedding prior to lambing and campylobacters were not detected in ewe faeces until day 3, after which their numbers increased for the next 12 d. In the second group, low numbers of campylobacters (433 g−1) were found in ewe faeces before lambing but did not increase until day 6 after lambing. By day 19 they reached 1·3 × 105g−1, the highest number found in sheep faeces during these studies.
When calculated on an annual basis, thermophilic campylobacters are shed in faeces of sheep at pasture at between a third and a half of the rate they are found in sheep intestines. These results confirm those found for the more limited sampling programme reported earlier ( Stanley et al. 1998a ). Carriage and shedding of campylobacters by sheep are, however, much higher than the 15% found by Cabrita et al. (1992) and the low levels suggested by Franco (1988). There is a slight difference in the shedding rate on different types of pasture, with shedding highest on saltmarsh grazing, followed by upland fell and then farm grazing. These differences may be related to the quality of grazing, with the poorer the grazing the poorer the digestibility ( Speedy 1980) and the greater the throughput of plant material and, hence, increased shedding. The differences in the rate of shedding on the different pastures are confirmed by the numbers of Campylobacter shed in faeces. It should be pointed out, however, that enumerations were only carried out in September in this study, and counts done on other occasions in the same habitats have not produced such high numbers. Numbers are therefore probably variable and are well below those found in sheep intestines ( Stanley et al. 1998a ).
The main species of Campylobacter isolated from faeces of sheep at pasture is Camp. jejuni, followed by Camp. coli and Camp. lari. This is the same pattern as found with campylobacters isolated from the intestines of sheep at slaughter ( Stanley et al. 1998a ).
Shedding of campylobacters by sheep has the potential to contaminate pastures and surface waters. For example, a sheep, grazing on farmland and excreting 1–2 kg faeces per day in September ( Spedding et al. 1981 ), would excrete between 3·5 × 107 and 7 × 107 campylobacters per day, and the 2000 sheep grazing the saltmarsh sampled in our study would excrete between 1·1 × 1011 and 2·2 × 1011 campylobacters per day. Run-off from sheep-grazed fells and farms in wet weather will contribute to the loading of campylobacters seen locally in groundwater, streams and rivers flowing through hills and farmland ( Bolton et al. 1987 ; Jones et al. 1990 ; Jones & Hobbs 1996; Stanley et al. 1998c ). Similarly, spring tides removing sheep faeces from the extensive saltmarshes bordering the Lune estuary and Morecambe Bay will contribute to the campylobacters detected in the estuary ( Zainuldin & Jones 1996) and the sea ( Jones et al. 1990 ; Obiri-Danso & Jones 1999). Sheep faeces also serve as a reservoir for the transfer of campylobacters by birds in the same way as has been demonstrated for bird-pecked cow pats ( Skirrow 1994). Similarly, the large numbers of campylobacters in sheep faeces on pasture and in stream water may lead to grazing sheep being continually challenged by campylobacters. The presence of campylobacters in 30% of rumen samples of sheep at slaughter suggests that ingestion of campylobacters with feed is a regular occurrence ( Stanley et al. 1998a ).
Although it is likely that intestinal infection is close to 100% ( Stanley et al. 1998a ), shedding of campylobacters in faeces varies considerably with the time of year. On some occasions, 100% of sheep are shedding campylobacters and on other occasions, none. Apart from the effects of lambing, the reasons for the peaks and troughs of shedding are speculative and are discussed below.
In the main study, ewes on farm pasture did not generally shed campylobacters during gestation (November through to March), but shedding resumed with the onset of lambing. We investigated the effects of lambing on shedding of campylobacters by ewes, and the colonization of new-born lambs with campylobacters, in two separate experiments. In the first, ewes were not shedding prior to lambing but began to shed 3 d after lambing, and in the second, where the ewes were shedding in low numbers prior to lambing, the number of campylobacters increased dramatically after lambing. Lambs born to the ewes shedding campylobacters prior to lambing became colonized more quickly than those born to the ewes which were not shedding prior to lambing. Such horizontal transfer of campylobacters from mothers and the immediate surroundings of the farm environment has also been shown for calves ( Stanley et al. 1998b ), pigs ( Weijtens et al. 1997 ) and turkeys ( Wallace et al. 1998 ). On the saltmarsh (May 1997), lambs were shedding campylobacters at a much higher rate than adults (adult and lamb faeces were not normally analysed separately in this study), in rather the same way that calves show a higher rate of infection with Campylobacter than adult cattle ( Grau 1991; Stanley et al. 1998b ).
On the fells and on the saltmarsh, shedding of Campylobacter increased markedly in September at the time when lambs were removed from their mothers. This fits in with the reports in the literature of outbreaks of scouring among weaning lambs, apparently due to Camp. jejuni ( Skirrow 1994).
The high shedding rate on the farm in May was associated with a change of pasture, and, similarly, the high rates of shedding on the saltmarsh in May 1996 and April 1997 coincided with the change from inland grazing onto the saltmarsh. On the fells, the September peak in shedding coincided with the movement of the sheep from moorland to rough pasture. Metabolic disorders are often associated with abrupt changes in diet or with movement from one place to another ( Hughes & Kershaw 1958; Spedding 1965). Abrupt changes in diet have been shown to cause increased shedding of Escherichia coli 0157:H7 by sheep ( Kudva et al. 1997 ). The high shedding rates on transfer to new pasture were also associated with the production of fluid stools. However, fluid stools are not always a sign of increased shedding of campylobacters. All the faecal samples taken from adult faeces on the three types of grazing in May 1997 were fluid but low in campylobacters. We have no explanation as to why shedding on the fields and on the fells in November and December should be so much higher than on the farm.
Scouring (diarrhoea), which normally lasts for only 24 h, is known to be associated with diseases, disorders and helminths ( Spedding 1965). The spring peak in shedding of Campylobacter coincides with the ‘spring rise’ in numerous parasitic larvae, which reach full development after lambing and at the onset of lactation ( Spedding 1965).
No association of shedding was associated with the stress caused by sheering or dipping. However, as scouring is thought to last only 24 h, it may be that the sampling programme missed any short-term effects.
Shedding of campylobacters by sheep was sometimes absent or very low. This happened with sheep on the farm during gestation but did not occur at the same time on the fells or the fields bordering the saltmarsh. On the farm, the low rates of shedding were also associated with supplementation of the diet. In temperate regions, pasture does not grow to any great extent from November to March and has to be supplemented with hay or silage. On the farm, this occurred when the sheep were housed inside from January to March. On the fells, it occurred in January when the pasture was covered in snow. Shedding was also low after the sheep received antibacterial vaccines in March. Although the vaccines target clostridia and Pasteurella, there may be an ‘immune boost’ which effects shedding of campylobacters.
The variability in shedding of campylobacters by sheep, shown by our work, illustrates that a programme of sampling needs to be carried out when testing sheep for shedding of bacteria. High or low values obtained on one or two occasions may not be typical. The work reported here reinforces the message of Stanley et al. (1998a) that the role of sheep in the environmental epidemiology of Campylobacter has been underestimated.
The authors are grateful for the co-operation and advice of the farmers involved in this study, to Susan Anderton and Kwasi Obiri-Danso who helped with sampling, and to MAFF (project CSA 2151) who supported the work on the colonization of lambs.
Present address: Department of Life Sciences, University of East London, Romford Road, London E15 4LZ, UK.