Effect of sampling method on the representative recovery of microorganisms from the surfaces of aquacultured finfish

The objective of this study was to determine if a gentle rinse procedure was equivalent to the combination of excision and homogenization with a stomacher for the relative removal of various microorganisms from finfish fillets. Fillets of hybrid striped bass and rainbow trout were obtained from local markets and sampled using three methods: rinse (R), excision followed by homogenization in a stomacher (S), and homogenization of fillets following a rinse (RS). Microorganisms were enumerated on selective and nonselective media, and randomly selected colonies from aerobic plate counts were identified using MIDI Sherlock and BIOLOG microbial identification systems. Enrichments and selective media were used for the isolation of Listeria monocytogenes, Salmonella spp., and Yersinia enterocolitica. This study confirms previous reports that stomaching is superior to rinsing for enumerating total microbial populations from fish fillets. Rinsing was more effective for rainbow trout than for striped bass. Sampling method did not affect the relative magnitude of plate counts on media selective for aeromonads, pseudomonads, Shewanella, lactic acid bacteria, enterics, and gram-positive cocci. In the compositional analysis of random isolates, R recovered significantly lower fractions of aeromonads than did S or RS, but sampling method did not affect the percent recovery of lactic acid bacteria, pseudomonads, Shewanella, Moraxellaceae, or Cytophaga/Flavobacterium. However, observations suggest that with increased replication, differences among Moraxellaceae, Pseudomonas, and gram positives might be significant. Only one L. monocytogenes colony was isolated, and no Salmonella or Y. enterocolitica, so the effect of sampling method could not be determined for these organisms. Differences in predominant bacterial populations were seen between fish species.