The critical importance of sample site in the determination of the frequency of oxytetracycline resistance in the effluent microflora of a fresh water fish farm

The frequency of oxytetracycline resistance in the microflora was studied in the effluent of a fresh water Atlantic salmon farm which had not used any oxytetracycline for three years. Resistance was defined as the ability to grow on casein peptone starch agar containing 25 mu g ml(-1) oxytetracycline. The farm effluent passed through a rolling drum filter and the retentate Row then entered a sedimentation tank. The effluent was sampled from the filter retentate as it left the filter and from the settled solids in the sedimentation tank. 12 independent samples were taken on four sampling days from two sampling sites. The frequency of resistance was dramatically different in samples taken at the two different locations in the farm outflow. In samples from the retentate of the outflow filter the median frequency of resistance was 0.47% (range 0.25-1.2%). In contrast in samples taken from a sedimentation tank downstream from the filter the median frequency was 38% (range 10-48%).Studies in oxytetracycline-free laboratory mesocosms demonstrated that in the presence of anaerobically decomposing fish feed, river sediments and river water the frequency of oxytetracycline resistant strains increased rapidly during incubation at 18 degrees C. After 14 days incubation the frequency had risen from 1.0% to 25%. In a similar mesocosm from which feed was omitted the frequency of resistance remained below 1%. It is suggested that elevated frequencies of oxytetracycline resistance may be encountered in environments where fish feed accumulates. These results have important implications for the design of surveys of the impact of antimicrobial agent use in fresh water fish farms.