Rodents and insectivores were trapped on ten organic farms (nine pig farms, one broiler farm) by using live-traps between August and October 2004. Farms were subdivided into five areas: feeding passage (near the feed trough), storage (inside the stable), outdoor area (next to the stable, solid/slatted concrete floor, sometimes roofed), pasture, and “nature area” (further from stable, wild vegetation). After CO₂ euthanasia, trapped animals were taken to the laboratory within hours. Upon arrival, their ceca and colons were processed.

Pig manure samples were collected on 6 of 10 farms in October 2004. Four farms were omitted because the livestock batch had meanwhile changed. On each farm, five fresh manure samples (10 g) were taken throughout the stables. These samples were pooled and analyzed for the presence of Salmonella and Campylobacter.

Rodent intestinal contents or pig manure were directly streaked on CCDA Campylobacter-selective agar plates (Oxoid CM739 + SR155) by using cotton swabs. These swabs were subsequently transferred to 5 ml of Preston enrichment broth (Oxoid CM67 + SR84 + SR204) without blood (8). Enrichment broth was incubated for 24 h at 41.5°C under microaerobic conditions and plated onto CCDA plates. All CCDA plates were incubated for 48 h at 41.5°C under microaerobic conditions and examined for presence of characteristic colonies of Campylobacter. Confirmation was performed by checking morphology and motility microscopically. The amplified fragment length polymorphism (AFLP) method was used for species identification and genotyping (3, 4).

For Salmonella detection, a swab containing 0.1 to 1 g of rodent intestinal content or pig manure was added to 9 ml of nonselective preenrichment medium, buffered peptone water (Biotrading K168B009). After preenrichment for 18 h at 37°C, 3 drops were transferred to Modified Semi-Solid Rappaport Vassiliadis agar plates (MSRV; Lab M, Lab150). MSRV plates were incubated at 41.5°C for 24 h. Growth on MSRV plates suspected to be Salmonella sp. was streaked on Brilliant Green Agar (Oxoid CM329) and examined for presence of suspected colonies after incubation for 18 to 24 h at 37°C. A suspected colony was then agglutinated with polyvalent anti-specific O antisera (Oxoid) and specific flagellar H antisera. Each isolate was biochemically confirmed (triple sugar iron [Oxoid] agar, urease, and lysine decarboxylase).

In total, 282 animals were examined, and 9 were positive for Campylobacter spp. (8 of 83 house mice, 1 of 8 brown rats). One Salmonella sp. strain Livingstone-positive house mouse was detected. Positive animals were only found in the storage, feeding passage and outdoor areas (Table 1). On three farms (A, B, and C), three single rodent isolates were obtained. On farm A, 1 of 6 (16.7%) house mice was positive (Campylobacter coli). On farm B, 1 of 8 (12.5%) brown rats caught was positive (C. coli). On farm C, 1 of 30 (3.3%) house mice was positive (Campylobacter spp.). This isolate was lost in follow-up and could not be speciated. On farm D, 1 C. hyointestinalis, 2 C. coli, and 3 C. jejuni strains were isolated from 6 of 15 (40%) house mice that were caught. The Salmonella sp. strain Livingstone-positive house mouse (also positive for C. jejuni) was caught here. Salmonella and Campylobacter spp. were not isolated from the pig manure at this farm. Overall, Campylobacter strains was isolated from pig manure on three of six farms (farms A, B, and E), but Salmonella strains were not encountered.

AFLP analysis revealed no direct shared genotypes of Campylobacter isolates from rodents and those from pigs at farms A and B (Fig. (Fig.1;1; a cutoff point of 95% similarity to define identical isolates was chosen). Campylobacter strains in two house mice caught at the same time at exactly the same trapping point on farm D were indistinguishable (Fig. (Fig.1,1, numbers 244 and 245). Further examination of this particular farm showed that almost all (five of six) Campylobacter-positive animals were found in one pig stable. The sixth infected animal was trapped in the feeding passage of another stable, only 25 meters away. No infections were encountered elsewhere.

Although we realize the limitations concerning the number of herds and their microbiological status, our study shows that on Dutch organic farms two rodent species were found to be carriers of Salmonella and Campylobacter strains. Salmonella prevalence was limited, a finding similar to a previous study (13) in which fecal pellets from house mice caught on-farm were analyzed (0 of 222 animals infected). However, Salmonella prevalence may increase when rodents are caught near a Salmonella sp.-positive herd or flock. The Campylobacter species found in rodents in the present study were C. jejuni, C. coli, and C. hyointestinalis subsp. hyointestinalis.

Although no shared Campylobacter genotypes in rodents and pigs were found, horizontal transmission by infected rodents cannot be excluded. Pigs can carry multiple Campylobacter strains (mixed infections), a phenomenon also observed in poultry (1, 7). In turn, this phenomenon may facilitate genetic exchange between different Campylobacter strains in the pig gastrointestinal tract under normal farm conditions. Such genetic instability under natural conditions could undermine the application of genetic subtyping (2). Hypothetically, infected rodents and pigs can acquire their infection from the same source or by infecting each other after having acquired the infection from some other source. For example, this could result from the consumption of an infected rodent by a pig or by a mouse foraging through infected pig manure.