Seven Swedish Yorkshire castrates from a conventional herd free from diseases according to the A-list of the International Office of Epizootics, and from Aujeszky's disease, atrophic rhinitis, Brachyspira spp, transmissible gastro-enteritis, porcine epidemic diarrhoea, porcine reproductive and respiratory syndrome and salmonellosis were used in this study.

The castrates had been weaned when they were 5 weeks old. They were between 10 and 11 weeks old (22–25 kg) when transferred to the experimental unit at the Department of Animal Nutrition and Management, Uppsala, Sweden. All pigs came from different litters. At the termination of the trial, the pigs weighed between 38.5 and 43.6 kg and were 14–15 weeks old.

The pigs were kept in separate pens with straw and the rooms were illuminated between 7:00 am and 7:00 pm every day. Each pen had a stone and a chain for the pigs to play with. A standard pig feed (Singel Flex, ODAL, Sweden) was offered twice daily at 7:30 am and 3:30 pm. The pigs were fed at a level of 4% of the mean live weight of the group. Water was available ad libitum. The health status of the animals was inspected at least twice daily and special attention was taken to keep the pigs washed and clean.

The pigs were acclimatised to the new environment for 7 days. Then (11 to 12 weeks of age; 27.5–33.0 kg body weight) they were surgically fitted with a Post Valve T-Caecum (PVTC) cannula as a preparation to a feeding trial. The construction and insertion of the cannula followed the procedure previously described by van Leeuwen et al. (1991). Lanolin-based zink oxide cream was used to avoid skin irritation around the PVTC-cannula, and the surgical stitches were removed after 9–11 days.

Prior to the surgery, the pigs were premedicated by one intramuscular injection with 75 mg each of tiletamin and zolazepasam (Zoletil forte vet., Virbac Laboratories, Carros, France) and 60 mg Azaperon (Stresnil™, Jansson & Cilag pharma, Wien, Austria). Within 30 min anaesthesia was induced and maintained by inhalation of O₂ and halothane. Post surgery the pigs were intramuscularily injected once with a long acting oxytetracycline (20 mg per kg body weight; Terramycin^® vet. Prolongatum, Pfizer, New York, USA), and they were also given an intramuscular injection of 0.3 mg buprenorphinum (Temgesic^®, Reckitt & Coleman, Hull, England) to reduce post-operative pain. Two of the pigs received one more injection of buprenorphinum on the day after surgery and 5 of the pigs received 2 more injections on the 2 days following surgery.

To document the enteric coliform flora before inserting the cannula, ileo-caecal and rectal samples were collected with a cotton swab during the surgery. To evaluate the influence of surgery and antimicrobial treatment, similar samples were collected 3, 7, 14 and 20 days post surgery.

All samples were spread on blood agar (blood agar base No.2; LabM, Salford, England + 5% horse blood) and incubated for 24 h at 37°C. After incubation, the bacterial growth were homogenized and dispersed in broth and frozen in -20°C until analysed. Then they were spread on McConkey agar and incubated for 24 h at 37°C. Twenty-four colony forming units from each sampling site and day were scrutinised by biochemical fingerprinting. Overall 1680 isolates were analysed using the Phene Plate (PhP) system (Ph Plate, Stockholm AB, Sweden). The isolates were inoculated on PhP-RS plates (Pheneplates^®, Biosys, Stockholm, Sweden), a system measuring the kinetics of bacterial growth in liquid medium in microtitre plates 1213. Each microtitre plate contains 11 dehydrated reagents, chosen to differentiate between coliforms 13. The metabolic response of each bacterial isolate to every substrate was measured at 620 nm using a microplate reader (Titertek Multiscan MCC/340, Labsystems OY, Helsinki, Finland) after 4, 7, 24 and 48 h of incubation at 37°C. The mean value of all readings was taken as the metabolic fingerprint for each isolate, and a dendrogram was constructed after pairwise comparison of biochemical fingerprints 8. Isolates showing higher similarities than the established identity level (97.5%) were regarded as identical and assigned to the same biochemical phenotype (BPT). The phenotypic diversity of the coliforms was measured as Simpsons index of diversity 6. The diversity is high (max value = 1) for a population comprising different BPTs and low (0) if only one BPT is present.

The effect of sampling day on diversities for the coliform populations at each sampling site were compared by the GLM procedure of SAS version 6.12 (SAS 1998). Pig and sampling day were used as main effects in the model. Results are presented as least square means with standard errors. Students t-test was used when means of the coliform diversity in the ileo-caecal ostium and in rectum were compared.

The pigs remained healthy during the convalescence period and showed no signs of disease following surgery. They ate all the feed they were offered and no feed refusals were observed.

The diversity of the coliform populations were studied in 24 isolates per pig and sampling site each day of sampling. In total 1680 isolates were scrutinised. The diversity was high before surgery, both at the ileo-caecal ostium and in rectum (Fig. 1). Following surgery and oxyte-tracycline treatment, no changes in the diversity of the enteric coliform populations were observed in any of the investigated spots (Fig. 1). Further, the diversity of the enteric coliform flora was similar at these sites of sampling (the ileo-caecal ostium and the rectum). Thus, no statistically significant differences within group over time, or between groups at the different sampling occasions were observed.