Fresh, iced Alaska pink salmon were stored upto 20days with samples of skin, gills and belly cavity analyzed for bacterial flora developments. Bacterial colonies isolated from plates used to estimate aerobic bacterial populations were identified using rapid cellular fatty acid analysis. Acidovorax spp. and Brevundimonas spp. were identified as part of the fresh fish microflora. The group consisting of Psychrobacter, Moraxella and Acinetobacter were present throughout salmon storage but were found only associated with fish skin and gills. Pseudomonas fluorescens and P. putida formed a major percentage of the spoilage flora and Shewanella putrefaciens comprised the minor portion in sampled tissues. The replacement of using the Sherlock Microbial Identification System for classical taxonomic tests resulted in time, material and labor savings. Also, the database provided a deeper view of the bacterial diversity and changes in bacterial flora that occurred during iced salmon storage.

Keywords: rapid identification, fatty acids, salmon, spoilage microflora

The most common method to determine the microbial load of fresh fish is the aerobic plate count and the spoilage threshold is 107 colony-forming units/g.¹ This value can vary during post-harvest storage of seafood and depends on the initial bacterial load, post-harvest handling, storage conditions, temperature, specific tissues sampled and composition of the bacterial flora.^2‒4 Although the microflora on fish is diverse, a small group of bacteria designated as “specific spoilage organisms” (SSO), namely Pseudomonas spp. and Shewanella putrefaciens, are able to cause deterioration and odor in iced fish.⁵ These SSO can interact with other bacteria to affect the dynamics in the fish microflora.⁶

The identification of bacteria involved in changes to the fish microflora can help in modeling the microbial community structure for particular seafood, assessing interactions within the community, predicting and extending shelf life of a product.⁵ The SSO load is estimated by using selective media but these media can support growth of non-target organisms giving an incorrect number of SSO.⁷ Hence, it would be advisable to identify the micro-organisms to the species level.

Identification of these micro-organisms by classical techniques is a time consuming, labor intensive and sometimes expensive process. The Sherlock® Microbial Identification System (MIS) is a fully automated system using direct computerized comparison of fatty acid analysis of test strains against the library database of >100,000 profiles to provide a rapid identification.^8,9 For seafood bacteria identification, the Sherlock® MIS has been limited to bacteria associated with aquaculture striped bass and biogenic amine-forming bacteria in canned anchovies.^10‒12 Recently, we determined that the Sherlock MIS could confirm and update the taxonomic classification of bacteria from our culture collection in a comparison study with strains from the American Type Culture Collection.¹³

The purpose of this study was to rapidly identify bacteria associated with the fresh salmon microflora found on iced fish skin, gills and belly cavity by using the Sherlock MIS and to follow the spoilage microflora developing during fish storage.

Fish processing

Fresh seine-caught pink salmon (Oncorhynchusgorbuscha) in the Gulf of Alaska and held in chilled seawater for <24h were procured during summer from a seafood processor in Kodiak, AK. Fish without physical damage were iced and brought to the Kodiak Seafood and Marine Science Centre’s pilot plant, de-iced, washed and maintained under iced condition till further processing. Fish were brought on two separate occasions and used for experiments to imitate commercial conditions of processing and handling of fillets.

Iced skin-on fillet storage

Within one hour, the first batch of fish (n=3) were headed, gutted, trimmed to remove the fins and washed briefly in tap water. Dressed fish were hand-filleted into boneless, skin-on fillets and rinsed with chilled tap water to remove bloodstains and slime. These fillets were placed individually in plastic bags closed with rubber bands to prevent contact with melting ice water. The bags were placed in a tote having a loose fitting lid on layer of flaked ice, then covered with more ice and kept in a walk-in cold room (4±1°C). The melting ice was replenished every 48h for 14days. The fish fillets had an average temperature maintained at 2±1°C. A random bag was pulled every threedays up to day15.

Gutted fish storage in ice

The second batch of pink salmon (n=4) was immediately degutted and rinsed thoroughly with tap water. Fish were placed belly down to prevent water accumulation and spaced apart on a layer of flaked ice, then covered with more ice. The totes were covered with lids and kept in the cold room. Melting ice was replenished every 24h for 19days. Initially and atdays 4, 6 and 20, one fish was taken for sampling.

Bacterial analysis

On each sampling day, the fish skin and belly cavity were swabbed (area of 10cm²) separately for conducting aerobic plate counts.¹⁴ Gills were aseptically cut and 25g of sample was macerated in 225mL of 0.1% (w/v) peptone water and serial dilutions were made in the diluents.¹⁵ Aliquots (0.1mL) of each serial dilution were spread-plated on duplicate plate count agar (PCA; Difco Lab, Detroit, MI) supplemented with 0.5% w/v NaCl. These plates were incubated at 25°C for 48-72h and then plates with 30-300 colonies were selected for counting.¹⁶ Total number of colonies counted from duplicate plates were averaged and used to calculate log colony-forming units (CFU) per cm² for skin and belly cavity and log CFU/g for gills.

From representative plates at each sampling period, ten colonies were randomly selected and restreaked to purity. These isolates were tested using the KOH reaction for cell wall type,¹⁷ catalase activity¹⁸ and oxidase test.¹⁹ Morphological characteristics (shape and motility) of the selected micro-organisms were observed under microscope (Montagesaez T-UL, 467065-9914, Zeiss, Germany) by using the 40X high-dry and 100X oil-immersion objectives. The pure cultures were collected individually from the PCA and stored in sterile 50% glycerol (Sigma Chemical Co., St. Louis, MO) and 40μL dimethylsulfoxide (Sigma) at -80°C until identification.

Frozen cultures were thawed and resuscitated in brain heart infusion (BBL, Sparks, MD) overnight at 25°C and then streaked on trypticase soy broth agar (30g of Trypticase soy broth (BBL) and 15g granulated agar (BBL) per liter) plates and incubated at 28°C for 24h. Fatty acid methyl esters were extracted and analyzed in duplicates using the procedure of Paisley²⁰ and the GC model 6850 (Agilent Technologies, Wilmington, DE) coupled with a flame ionization detector. The GC conditions were monitored by the Sherlock® MIS software (Microbial ID Inc., Newark, DE). The GC was calibrated using a calibration standard (MIDI Part No. 1300-AA) for the Sherlock® Rapid Method.²⁰ During all the calibration mixture analysis, the peak percent named for the standard was >99% with the root mean square error <0.003. Strains with Similarity Index (SI)>0.5 with a separation of >0.1 between first and second ranks are considered as good library comparisons. If a closely-related genus or species was listed <0.1SI, the identification of that particular isolate was reported as a combination.

The APC on the skin was <4 log CFU/cm² for the first week then reached 6.6log CFU/cm² at 15days of iced storage (Figure 1). Low initial counts in the belly cavity (Figure 2) indicated that the intestines were intact after capture till the fish were gutted and did not contaminate the belly cavity. Low bacterial counts could be due to the washing effect during on-board chilled seawater storage, evisceration and sample rinsing prior to the start of the experiment.²¹Additionally, the washing effect of melting ice and its replacement with fresh ice every 24h could have also led to low APC during initial iced storage. Subsequent growth of bacteria on the skin surface, gills and belly cavity was attributed to the psychrotrophic or psychrotolerant nature of bacteria adapted to chill temperatures.²² Although the initial count of belly was approximately similar to the skin, the APC reached approximately 8 log CFU/cm² while gills with a higher initial APC count reached approximately 7log CFU/g, at 20days of storage (Figure 3).

Of 60 isolates from skin-on salmon fillets collected over 15days of iced storage, 93% were Gram-negative and the remainder was Gram-positive. The micro flora changed from very diverse on day 0 to comprising three genera by day 15 (Figure 1). From the 104 bacterial isolates collected from the belly cavity and gills, 97% were Gram-negative and the remainder was Gram-positive. These trends were similar to the ones observed on salmon skin fillets indicating that the majority of spoilage microflora on pink salmon belongs to Gram-negative group of bacteria.²¹

Identification of bacterial isolates indicated a presence of some under reported seafood bacterial species such as Acidovorax, Delftia and Brevundimonas during the early part of iced salmon storage. Nedoluha and Westhoff^10,23,24 indicated the presence of these genera identified using Sherlock® MIS in aqua cultured striped bas.^10,23,24 The formation of Acidovorax²⁵ and Delftia²⁶ and their low presence during the first fewdays of iced fish storage may indicate their non-importance in fish spoilage as compared to Pseudomonas spp. and S. putrefaciens. Brevundimonas vesicularis and B. diminutaare delineated from Pseudomonas.²⁷ Although Brevundimonas spp. are mainly associated with clinical specimens,²⁸ B. vesicularis and B. diminuta have been isolated from seafood.^6,29,30

Bacteria belonging to genus Flavobacterium were identified in the spoilage microflora of pink salmon along with Chryseobacterium³¹ and Elizabethkingia.³² These bacteria have been reported only as Flavobacterium spp. which provides no information about speciation. Isolates from pink salmon identified using Sherlock® MIS as F. johnsoniae, C. indologenes and C. meningo septicum have been reported by other researchers.^10,23,24,33

The seafood bacteria grouping of Psychrobacter, Moraxella and Acinetobacter were isolated from iced pink salmon and were confirmed using the Sherlock® MIS and comprised 20-50% of the skin microflora. Species-level overlap was due to a very low SI and <0.1 SI difference between first ranks of identifications. Presence of these bacteria throughout storage irrespective of sampling regimen could indicate their commensal role in pink salmon spoilage development. As the ice storage progressed, the percentage of the microflora was maintained, then decreased possibly due to competitive exclusion by motile Pseudomonas spp. during the later stage of iced storage, similar to observations by McMeekin TA.³⁴

Primarily P. fluorescens - P. putida (40-80%) and S. putrefaciens (10%) were identified at the three sampling locations of pink salmon on the final day of iced storage (Figures 1‒3). The occurrence of P. fluorescens and P. putida increased with a corresponding decrease in the variety of bacterial species present during storage. The predominance of these organisms can be attributed to the ability of Pseudomonas to grow rapidly at chill temperature³⁵ and suppressed growth of other bacteria.^36,37 The importance of Pseudomonas spp. as SSO can be attributed to the ability of psychrotrophic Pseudomonas strains that exhibit lipolytic and proteolytic activity resulting in fish spoilage.^38,39 Pseudomonas putida is recognized as a less important fish spoilage bacterium due to a lack of proteolytic activity⁴⁰ but isolates from Mediterranean hake produced putrescine.⁴¹

The bacterial flora observed on gills and skin for day 0 of pink salmon was typically a diverse microflora and expected of fresh fish harvested from Alaskan waters. As storage progressed, there was a change in the microflora, as observed in a previous report.⁴² The presence of P. fluorescens or P. putida at 80% of the total microflora on final day of ice storage in all the tissues reinforces the importance of Pseudomonas spp. in the spoilage process⁴³ of pink salmon stored in ice. Presence of P. immobilis and M. catarrhalis was observed throughout the storage period and their commensal contribution to the spoilage process needs further research.In conclusion, the Sherlock® MIS was useful in identifying psychrotrophic seafood bacteria to the species level and can be used as a rapid method for assessing seafood quality and estimating remaining shelf life

Figure 1 Bacterial flora changes in skin of Alaska pink salmon during 15days in iced storage.

Figure 2 Bacterial flora changes in belly cavity of Alaska pink salmon during 20days in iced storage.

Figure 3 Bacterial flora changes in gills of Alaska pink salmon during 20days in iced storage.

This project was funded through USDA-CSREES grant #2004-34404-15017, the Alaska Sea Grant with funds from the National Oceanic and Atmospheric Administration Office of Sea Grant, Department of Commerce, under grant no. NA06OAR4170013 (project no. E/142-01) and from the University of Alaska Fairbanks with funds appropriated by the State.

Author declares that there is no conflict of interest.

1. International Commission on Microbiological Specifications for Foods (ICMSF). Sampling plans for fish and fishery products. In: M Ingram, DF Bray, DS Clark, editors. Microorganisms in Foods Sampling for Microbiological Analysis. Principles and Specific Applications. Toronto, Canada: University of Toronto Press. 1978. p. 92‒104.
2. Ola JB, Oladipo AE. Storage life of croaker (Pseudotholitus senegalensis) in ice and ambient temperature. African J Biomed Res. 2004;7:13‒17.
3. Jeyasekaran G, Maheshwari K, Ganesan P, et al. Quality changes in ice-stored tropical wire-netting reef cod (Epinephelus Merra) J Food Process Preserv. 2005;29(2):165‒182.
4. Hozbor MC, Saiz AI, Yeannes MI, et al. Microbiological changes and its correlation with quality indices during aerobic iced storage of sea salmon (Pseudopercis semifasciata). Food Sci Technol. 2006;39(2):99‒104.
5. Gram L, Dalgaard P. Fish spoilage bacteria-problems and solutions. Curr Opin Biotechnol. 2002;13(3):262‒266.
6. Baixas-Nogueras S, Bover-Cid S, Veciana-Nogues MT, et al. Suitability of volatile amines as freshness indexes for iced Mediterranean hake. J Food Sci. 2003;68(5):1607‒1610.
7. Salvat G, Rudelle S, Humbert F, et al. A selective medium for the rapid detection by an impedance technique of Pseudomonas spp. associated with poultry meat. J Appl Microbiol. 1997;83(4):456‒463.
8. O’Hara CM. Manual and automated instrumentation for identification of Enterobacteriaceae and other aerobic gram-negative bacilli. Clin Microbiol Rev. 2005;18(1):147‒162.
9. Himelbloom BH, ACM Oliveira, TS Shetty. Rapid methods for the identification of seafood micro-organisms. In: C Alasalvar, F Shahidi, K Miyashita, editors. Handbook of Seafood Quality, Safety, and Health Applications. Oxford: Blackwell Publishing; 2010. p. 226‒236.
10. Nedoluha PC, Westhoff D. Microbiological analysis of striped bass Morone saxatilis grown in flow-through tanks. J Food Protect. 1995;58(12):1363‒1368.
11. Kim SH, Eun JB, Chen TY, et al. Evaluation of histamine and other biogenic amines and bacterial isolation in canned anchovies recalled by the USFDA. J Food Sci. 2004;69(6):M157‒M162.
12. Lee H, Kim SH, Sang CI, et al. Histamine and other biogenic amines and bacterial isolation in retail canned anchovies. J Food Sci. 2005;70(2):C145‒C150.
13. Morey A, Oliveira ACM, Himelbloom BH. Identification of seafood bacteria from cellular fatty acid analysis via the Sherlock® Microbial Identification System. J Biol Life Sci. 2013;4(2):139‒153.
14. Tretsven WI. Bacteriological survey of filleting processes in the Pacific Northwest. II. Swab technique for bacteriological sampling. J Milk Food Technol. 1963;26:383‒388.
15. Crapo C, Himelbloom B, Vitt S, et al. Ozone efficacy as a bactericide in seafood processing. J Aquatic Food Prod Technol. 2004;13(1):111‒123.
16. Harrigan WF. Determination of the number and detection of viable microorganisms in a sample. In: Harrigan WF, editor. Laboratory Methods in Food Microbiology. San Diego, CA: Academic Press. 1998. p. 52‒70.
17. Powers EM. Efficiency of the Ryu nonstaining KOH technique for rapidly determining gram reactions of food-borne and waterborne bacteria and yeasts. Appl Environ Microbiol. 1995;61(10):3756‒3758.
18. Harrigan WF. Biochemical tests for identification of organisms. In: Harrigan WF, editor. Laboratory Methods in Food Microbiology. San Diego, CA: Academic Press. 1998. p. 100‒118.
19. Kovacs N. Identification of Pseudomonas pyocyanea by the oxidase reaction. Nature. 1956;178(4535):703.
20. Paisley R. Sample preparation, standard TSBA40 method. In: Training manual MIS whole cell fatty acid analysis by gas chromatography. Newark, DE: MIDI Inc; 2004. p. D5‒D21.
21. Himelbloom BH, Crapo C, Brown EK, et al. Pink salmon (Oncorhynchus Gorbuscha) quality during ice and chilled seawater storage. J Food Qual. 1994;17(3):197‒210.
22. Gram L, HH Huss. Fresh and processed fish and shellfish. In: BM Lund, AC Baird-Parker, GW Gould, editors. The Microbiological Safety and Quality of Foods. London: Chapman & Hall; 2000. p. 472‒506.
23. Nedoluha PC, Westhoff D. Microbiology of striped bass grown in three aquaculture systems. Food Microbiol. 1997;14(3):255‒264.
24. Nedoluha PC, Westhoff D. Microbiological analysis of striped bass (Morone saxatilis) grown in a recirculating system. J Food Protect. 1997;60(8):948‒953.
25. Willems A, Falsen E, Pot B, et al. Acidovorax, a new genus for Pseudomonas facilis, Pseudomonas delafieldii, E. Flasen (EF) group 13, EF group 16, and several clinical isolates, with the species Acidovoraxfacilis comb. nov., Acidovorax delafieldii comb. nov., and Acidovorax temperans sp. nov. Int J Syst Bacteriol. 1990;40(4):384‒398.
26. Wen A, Fegan M, Hayward C, et al. Phylogenetic relationships among members of the Comamonadaceae, and description of Delftia acidovorans (den Dooren de Jong 1926 and Tamaoka et al. 1987) gen. nov., comb. nov. Int J Syst Bacteriol. 1999;49(2):567‒576.
27. Segers P, Vancanneyt M, Pot B, et al. Classification of Pseudomonas diminuta Leifson and Hugh 1954 and Pseudomonas vesicularis Busing, Doll, and Freytag 1953 in Brevundimonas gen. nov. as Brevundimonas diminuta comb. nov. and Brevundimonas vesicularis comb. nov., respectively. Int J Syst Bacteriol. 1994;44(3):499‒510.
28. Vancanneyt M, P Segers, WR Abraham, et al. Genus III. Brevundimonas Segers, Vancanneyt, Pot, Torck, Hoste, Dewettinck, Falsen, Kersters and De Vos 1994, 507VP emend. Abraham, Strompl, Meyer, Lindholst, Moore, Christ, Vancanneyt, Tindall, Bennasar, Smit and Tesar 1999, 1070. In: Brenner J, NR Krieg, JT Stanley, editors. Bergey’s Manual of Systematic Bacteriology. 2nd ed. New York: Springer; 2005. p. 308‒316.
29. Rodriguez O, J Barros-Velazquez, A Ojea, et al. Effect of chilled storage in flow ice on the microbial quality and shelf life of farmed turbot (Psetta maxima). Isolation and identification of major proteolytic bacteria. In: Proceedings of the First Joint Trans-Atlantic Fisheries Technology Conference (TAFT 2003), Icelandic Fisheries Laboratories, Reykjavik, Iceland; 2003. p. 73‒74.
30. Kapetanovic D, Kurtovic B, Vardic I, et al. Preliminary studies on bacterial diversity of cultured bluefin tuna Thunnus thynnus from the Adriatic Sea. Aquacult Res. 2006;37(12):1265‒1266.
31. Vandamme P, Bernardet JF, Segers P, et al. New perspectives in the classification of the flavobacteria: Description of Chryseobacterium gen. nov., Bergeyella gen. nov., and Empedobacter nom. rev. Int J Syst Bacteriol. 1994;44(4):827‒831.
32. Kim KK, Kim MK, Lim JH, et al. Transfer of Chryseobacterium meningosepticum and Chryseobacterium miricola to Elizabethkingia gen. nov. as Elizabethkingia meningoseptica comb. nov. and Elizabethkingia miricola comb. nov. Int J Syst Evol Microbiol. 2005;55:1287‒1293.
33. Osterhout GJ, Shull VH, Dick JD. Identification of clinical isolates of gram-negative nonfermentative bacteria by an automated cellular fatty acid identification system. J Clin Microbiol. 1991;29(9):1822‒1830.
34. McMeekin TA. Spoilage association of chicken leg muscle. Appl Environ Microbiol. 1977;33(6):1244‒1246.
35. Massa AE, Palacios DL, Paredi ME, et al. Postmortem changes in quality indices of ice-stored flounder (Paralichthys patagonicus). J Food Biochem. 2005;29(5):570‒590.
36. Gram L. Inhibitory effect against pathogenic and spoilage bacteria of Pseudomonas strains isolated from spoiled and fresh fish. Appl Environ Microbiol. 1993;59(7):2197‒2203.
37. Gram L. Siderophore-mediated iron sequestering by Shewanella putrefaciens. Appl Environ Microbiol. 1994;60(6):2132‒2136.
38. Surette M, Gill T, MacLean S. Purification and characterization of purine nucleoside phosphorylase from Proteus vulgaris. Appl Environ Microbiol. 1990;56(5):1435‒1439.
39. Olafsdottir G, Lauzon HL, Martinsdottir E, et al. Influence of storage temperature on microbial spoilage characteristics of haddock fillets (Melanogrammus aeglefinus) evaluated by multivariate quality prediction. Int J Food Microbiol. 2006;111(2): 112-125.
40. Gennari M, Dragotto F. A study of the incidence of different fluorescent Pseudomonas species and biovars in the microflora of fresh and spoiled meat and fish, raw milk, cheese, soil and water. J Appl Bacteriol. 1992;72(4):281‒288.
41. Baixas-Nogueras S, Bover-Sid S, Veciana-Nogues MT, et al. Amino acid-decarboxylase activity in bacteria associated with Mediterranean hake spoilage. Eur Food Res Technol. 2003;217(2):164‒167.
42. Crapo C, Himelbloom B. Spoilage and histamine in whole Pacific herring (Clupea harengus Pallasi) and pink salmon (Oncorhynchus gorbuscha) fillets. J Food Safety. 1999;19(1):45‒55.
43. Huss HH. Post mortem changes in fish. In: Quality and Quality Changes in Fresh Fish, Food Agricult Org of the UN. Rome, Italy; 1995. 348p.