Fish and seafood are an important food source for a large section of the world’s population, more so in developing countries 1. Unfortunately, seafood still accounts for a considerable number of outbreaks of food-borne illness 2.
Since Salmonella was first isolated in 1880, it has been of significant interest to food microbiologists as one of the leading causes of food-borne bacterial infections. Despite extensive public health measures over the past century, it remains one of the most commonly identified causes of bacterial food-borne disease in the developed countries and a significant cause of morbidity and mortality in the developing world 3.
Most fish and seafood are now exposed to animal or human wastes from the local environment as a result of dumping of untreated, contaminated wastes into the marine environment and can easily be contaminated by a wide variety of pathogenic organisms including Salmonella. Contamination with Salmonella can also result during harvest and afterwards during food handling and processing 4.
According to the European Food Safety Authority, Salmonella is still the main causative agent (35.4%) of all reported food-borne outbreaks. Fish and fish products accounted for 1.4% of these outbreaks whereas crustaceans, shellfish, and mollusks accounted for 1%. In the USA, between 1973 and 2006, Salmonella was responsible for 18 outbreaks associated with seafood, resulting in 374 cases of Salmonellosis 5,6.
Salmonella-contaminated seafood usually looks and smells normal. It is therefore essential that every effort is made by all those in the seafood production and supply industries to exclude the organism from seafood wherever possible 7.
The methods for the detection of Salmonella in food have been studied intensely for many decades. This has led to a number of analytical procedures for their isolation and identification including a variety of conventional and non conventional methods 8.
Conventional culture methods for the recovery of Salmonella from foods involve pre-enrichment of the food sample in nonselective media, followed by enrichment in a selective enrichment medium, streaking on selective agar medium, and then confirmation of isolates biochemically and serologically 9. These methods are laborious, intensive, and time-consuming 10,11.
Since 1990, a range of chromogenic media have been developed for the detection of Salmonella spp. The inclusion of chromogenic enzyme substrates that release colored dyes upon hydrolysis by microbial enzymes facilitates the differentiation of polymicrobial cultures, enabling such media to target pathogens with high specificity 12,13.
Most of these media have been tested for isolation of Salmonella from clinical samples. Little is known about the performance of Chromagar Salmonella Plus (CASP) in isolation of Salmonella from food samples, especially seafood. This study was carried out to determine the percentage of Salmonella in some Egyptian seafood sold in Alexandria markets and to assess the validity of Chromagar medium for the detection of Salmonella in some seafood.
Materials and methods
This study was carried out during the period from October 2009 until December 2009. A total of 225 seafood samples were collected from 11 localities in Alexandria City (Egypt). (In a pilot study, four of 24 seafood samples (16.6%) were contaminated with pathogenic organisms; using precision 5 at 95% CI, a minimum required sample size was calculated to be 213 using EPI info 6, version 2003, Atlanta, Georgia.) Seventy-five samples of each of shrimp, gandofli, and river mussel (om-elkhloul) were collected, each in a sterile container, labeled, and transferred, in an icebox, to the laboratory for examination within 1–2 h. The samples were scrubbed, rinsed with tap water, and then gandofli and river mussel (om-elkhloul) were opened aseptically and the flesh was collected. Twenty-five grams of each sample was homogenized in a stomacher for 1 min, and then the samples were pre-enriched in 225 ml of buffered peptone broth (Biolife, Italy) and incubated for 18–20 h at 37°C 14. A volume of 0.1 ml of the pre-enriched sample was transferred to each of Rappaport Vassiliadis (RV) broth (Biolife) and 10 ml of tetrathionate (TT) broth (Biolife); both media were incubated at 42°C for 24 h 15. A loopful from a well-shaken selective enrichment broth was streaked onto each of CASP (CHROMagar, Paris, France), xylose lysine desoxycholate (XLD) (Oxoid, Hampshire, UK), and Salmonella–Shigella (SS) (Difco, France) agar plates.
All plates were incubated aerobically at 37°C for 24 h and then examined for typical Salmonella colonies 16. According to the manufacturer, colonies of Salmonella appeared as mauve colonies on CASP medium. Enterobacteriaceae other than Salmonella appeared as blue (as Escherichia coli) and colorless colonies (as Proteus spp.). On XLD agar, Salmonella appeared as shiny pink colonies with or without black centers and sometimes they appeared as glossy large colonies 3–5 mm in diameter. On SS agar, the colonies appear as clear, colorless, and transparent and may have black centers because of H2S production 17. Biochemically identified Salmonella isolates were confirmed serologically using polyvalent Salmonella O antiserum poly A–I & Vi (Difco) 15. False-positive colonies were identified biochemically; they included Citrobacter freundii, Proteus spp., E. coli, and Pseudomonas aeruginosa18.
The data were collected and statistical analysis was carried out manually. The statistical tests determined sensitivity, specificity, positive predictive value, negative predictive value, and accuracy 19. A false-positive result was considered when a typical-appearing Salmonella colony was not identified as Salmonella by biochemical and serological tests. A false-negative result was considered for a specific agar medium when this particular medium did not recover Salmonella whereas another agar medium recovered it. The gold standard against which the results were compared was the total number of colonies identified as true Salmonella colonies, by biochemical and serological reactions, for all the combinations of enrichment broth and culture media used.
The present study showed that out of the 225 seafood samples examined, 22 (9.8%) samples were contaminated with Salmonella. Salmonella was detected in 17.3, 8, and 4% of shrimp, gandofli, and river mussel (om-elkhloul), respectively (Fig. 1). This difference was statistically significant at P less than or equal to 0.05 (χ2=7.96, P=0.019).
The highest percentage of Salmonella isolation was from the samples collected from Karmouz (15%), followed by Khourshid (12.5%), EL-Manshia and Mahattet-Misr (12%), and then EL-Ebrahemia, EL-Asafra, and Sedi-Bishr (10% each), followed by Bahary (6.7%). Salmonella was not detected from the seafood samples collected from EL-Mandara, EL-Hadara, and Moharram-Beik (Fig. 2).
When RV was used as enrichment, CASP medium detected 21 (95.5%) of 22 positive Salmonella samples, with no false-positive results. These results were associated with a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 95.5, 100, 100, 99.5, and 99.5%, respectively. κ was 0.974 (SE of κ=0.026); the strength of agreement is considered to be very good with the total number of Salmonella identified in all samples (Table 1).
With the XLD medium, 16 (72.7%) of the 22 Salmonella positive samples could be detected, with no false-positive results. These results yielded a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 72.7, 100, 100, 97.1, and 97.3%, respectively. κ was 0.828 (SE of κ=0.068); the strength of agreement was also considered to be very good (Table 2).
With SS agar, 31 isolates were detected that presumably appeared as Salmonella. Only 18 isolates (58.1%) were true positive following both biochemical and serological confirmation whereas the remaining 13 isolates (41.9%) proved to be false positive. These figures were associated with a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 81.8, 93.6, 58.1, 97.9, and 92.4%, respectively. κ was 0.638 (SE of κ=0.080); the strength of agreement was ranked as good (Table 3).
Taking into account all these results to determine the total number of Salmonella detected, we can make a comparison of our results. The highest sensitivity was obtained using CASP medium (95.5%), followed by SS agar (81.8%), whereas the least sensitivity was obtained using XLD agar (72.7%). In terms of specificity, both CASP medium and XLD agar yielded the same specificity (100%) whereas that of SS agar was (93.6%). Moreover, the accuracy of CASP medium was (99.5%), followed by XLD agar (97.3%) whereas the least accuracy was obtained using SS agar (92.4%). CASP medium and XLD yielded a very good agreement with the calculated data considered as the gold standard, whereas SS was only in good agreement.
When TT broth was used as enrichment, CASP medium isolated 18 (81.8%) Salmonella isolates, with no false-positive results. These results were associated with a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 81.8, 99.5, 94.7, 98, and 97.8%, respectively. κ was 0.866 (SE of κ=0.059); the strength of agreement was very good (Table 1).
XLD agar detected 42 isolates that presumably appeared as Salmonella. Only 11 (26.8%) were true positive following both biochemical and serological confirmation, whereas 31 (73.8%) isolates were proved to be false positive. These results yielded a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 50, 84.7, 26.19, 60.01, and 81.3%, respectively. κ was 0.247 (SE of κ=0.081); unlike the results with RV; the strength of agreement was calculated to be fair (Table 2).
SS agar detected 136 isolates that presumably appeared as Salmonella. Only 14 isolates (10.3%) were true positive following both biochemical and serological confirmation, whereas 122 isolates (89.7%) were false positive. These results were associated with a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 63.6, 40, 10.3, 91, and 42.2%, respectively. It also yielded poor strength of agreement with the total number of colonies identified (Table 3).
These results shows that when TT broth was used as enrichment, the highest sensitivity obtained was with CASP medium (81.8%), followed by SS agar (63.6%), whereas the least sensitivity was with XLD agar (50%). In terms of specificity, CASP medium yielded a specificity of (99.5%), followed by XLD agar (84.7%), whereas the least specificity was that of SS agar (40%). Moreover, the accuracy of CASP medium was (97.8%), followed by XLD agar (81.3%), whereas the least accuracy was that of SS agar (42.2%). Only CASP showed a good agreement with TT broth.
Contamination of seafood with Salmonella is a major public health concern. The presence of Salmonella spp. in seafood has been reported in Vietnam, India, Morocco, China, and Japan 4.
In the present study, Salmonella was recovered from 9.8% of 225 seafood samples, 17.3% of shrimp samples, 8% of gandofli samples, and 4% of river mussel (om-elkhloul) samples. Almost similar results were obtained by Heinitz et al. 20 and Brands et al. 21 in the USA, who noted an overall prevalence of Salmonella of 7.2% for imported seafood and 7.4% for oysters, respectively 20,21. Higher results were reported by Kumar et al. 22; in India, they detected Salmonella in 34.2% of clam, 31% of mussel, 26% of shrimp samples, 12.5% of oyster, and 9.6% of crab samples. In contrast, lower results were obtained by Heinitz et al. 20, who reported Salmonella in 1.3% of domestic seafood in the USA.
It should be emphasized that in the present study, the highest prevalence of Salmonella contamination (17.3%) was obtained from shrimp; this can be attributed to the fact that shrimp obtained from culture environments were studied. Some studies investigating contamination of shrimp in a culture environment suggested that animal manure and contaminated feeds added to grow-out ponds are possible sources of Salmonella whereas others have claimed that sediment and water are possible sources of contamination 23.
In the present study, the least percent of Salmonella was obtained from river mussel (om-elkhloul) samples. This could be attributed to the fact that river mussel are cleaned from dirt and plankton and salted by the addition of NaCl before they are sold. Almost similar results were obtained by Mansour et al. 24, who suggested that the risk of food poisoning because of Salmonella from consumption of river mussel (om-elkhloul) is reduced by the addition of salt and lemon before consumption.
Oysters pose a particular problem because they are filter feeders; therefore, they are ideally suited to trap all bacteria and viruses, pathogenic or otherwise, that live in the waters. This might explain the presence of Salmonella in gandofli (8%) 25.
In the present study, in terms of the localities from which different seafood samples were obtained, the highest percent of Salmonella was from the samples collected from Karmouz (15%), followed by Khourshid (12.5%), EL-Manshia and Mahattet-Misr (12%), and then EL-Ebrahemia, EL-Asafra, and Sedi-Bishr (10%), followed by Bahary (6.7%). Improper cleaning of storage and preparation areas and unclean utensils were observed in markets of Karmouz, Khourshid, EL-Manshia, Mahattet-Misr, EL-Ebrahemia, EL-Asafra, and Sedi-Bishr. Several studies have reported the occurrence of Salmonella in wholesale markets, where they isolated Salmonella from utensils (2%) and floor swab samples (4%) 26,27. Fresh seafood was rapidly sold in Bahary to all localities of Alexandria; this may explain why the percentage of Salmonella in Bahary was less than that in the other localities. However, none of the seafood samples collected from EL-Mandara, EL-Hadara, and Moharram-Beik showed any Salmonella. It was observed that seafood was kept on a thick bed of fresh ice as a method of preservation at these localities.
Contamination of food with Salmonella is a major public health concern; this has led to a number of different analytical procedures for their isolation and identification 7,8. Selective enrichment broths are used to favor the growth of Salmonella spp. to be detectable by plating and to restrict the growth of other competing organisms. The US Food and Drug Administration method recommends selective enrichment in RV broth and TT for Salmonella isolation from fish and fish products 15.
In the present study, the retrieval of Salmonella colonies from the same sample was more by RV than by TT on all the media used, although not statistically significant (at P≤0.05). Compared with TT broth, RV had a higher sensitivity with CASP (95.5 vs. 81.8%), XLD (72.7 vs. 50%), and SS (81.8 vs. 63.6%) and higher specificity with CASP (100 vs. 99.5%), XLD (100 vs. 84.7%), and SS (93.6 vs. 40%). Several authors have observed that RV is the most effective enrichment broth for Salmonella28,29.
Rybolt et al. 30,31 reported that TT, containing calcium carbonate, provides an optimal environment for Salmonella to proliferate, but at the same time, other microorganisms that are present may grow. Therefore, the use of RV, which contains malachite green, a substance that is toxic to many bacterial species, eliminates the competing organisms when this secondary enrichment is used. This may explain the better isolation rate of RV.
Selective plating is an essential step in the isolation and differentiation of Salmonella species from other members of the family Enterobacteriaceae. There is a wide range of commercially selective differential agars to isolate Salmonella17. In the present study, CASP was the most efficient medium as it had the highest recovery rate; it also yielded the highest sensitivity, specificity, and accuracy with both RV and TT broth. This could be attributed to the fact that CASP yielded the least number of false-positive and false-negative results.
Rall et al. 16 tested three classic plating media, SS, brilliant green agar, and XLD, and two chromogenic agars, Rambach and CHROMagar Salmonella. Among 100 poultry carcasses, the chromogenic media yielded better results compared with classic ones, with less false-positive results. The most effective isolation medium was CHROMagar Salmonella, where Salmonella was identified in 79.3% of the positive samples, followed by Rambach (48%). Positivity for Salmonella using classic media was 13.8% for brilliant green agar, 27.6% for SS, and 34.5% for XLD.
In many studies, chromogenic media show a proven advantage over conventional culture media because of a superior detection rate for target pathogens or a superior differentiation of mixed cultures 32,33. In contrast, Perez et al. 34 found that the CHROMagar Salmonella was less sensitive than Hektoen agar. Also, Baraheem et al.35 reported that CHROMagar Salmonella was less specific than XLD, as it yielded a specificity of 69%, whereas XLD yielded a specificity of 77.6%.
In the present study, CASP medium yielded only one false-positive result (P. aeruginosa) when TT was used for enrichment. Rall et al. 16 reported that the lowest number of false-positive colonies was observed with CHROMagar, and they were mainly P. aeruginosa or other small Gram-negative rods. Maddocks et al. 36 and Gaillot et al. 37 also observed that P. aeruginosa, Aeromonas hydrophila and Candida albicans can grow on this medium and can be ruled out from Salmonella spp. in a few minutes with a minimum number of rapid and inexpensive tests (colony morphology, oxidase test, or direct microscopy). Eigner et al. 12 tested a modified CHROMagar Salmonella formula containing cefsulodin and amphotericin B and found that Pseudomonads and yeasts were effectively inhibited.
In the present study, no false-positive results were obtained with XLD when samples were enriched in RV, but it yielded 73.8% false-positive results when enriched in TT. In contrast, SS yielded 41.93 and 89.7% false-positive results when enriched in RV and TT, respectively.
Maddocks et al. 36 reported that the combination of SS and XLD resulted in 43% false-positive results whereas CHROMagar Salmonella was associated with 15% false-positive results. False-positive results on XLD and SS agar media in the present study were mainly Proteus, Citrobacter, and P. aeruginosa. As Proteus and Citrobacter are commonly found in foods, the number of false-positive colonies in classic media may be very large. Wray and Wray 17 reported that the major disadvantage of XLD and SS is the high false-positive rate primarily because of Proteus spp. Rall et al. 16 reported that in classic media, Salmonella, Proteus, and Citrobacter present similar biotypes, they give a negative test for lactose fermentation, and they produce H2S. Gaillot et al. 37 reported that such false-positive organisms can be routinely differentiated from Salmonella spp. by combinations of various techniques that require at least 24–48 h of incubation. In our study, the use of CASP medium for the detection of Salmonella from seafood samples showed high levels of sensitivity and specificity, easy differentiation of Salmonella from other bacteria, and reduced the time to final identification of Salmonella spp. by 24–48 h, resulting in considerable cost savings. Others, especially SS agar, were associated with tedious work and waste of both time and effort in differentiation between Salmonella and large number of false-positive colonies.
Conclusion and recommendation
Seafood is a potential vehicle for transmission of food-borne diseases. RV was found to be superior to TT in the isolation of Salmonella from seafood samples. CASP medium was more accurate than XLD and SS in the detection of Salmonella from seafood samples; it provides a time-saving method for the detection and presumptive identification of Salmonella from seafood specimens. Interpretation of colors was easy, and all colonies of Salmonella tested showed the same color and morphology. We recommend that CASP medium should be tested against more Salmonella-positive samples before we recommend its use as a screening plating medium for Salmonella in seafood as it allowed rapid and easy visual detection of Salmonella colonies.
The authors thank CHROMAagar (Paris, France) for their generous supply of the CHROMagar Plus medium.
Conflicts of interest
There are no conflicts of interest.
1. Morton S.Foodalert!: the ultimate sourcebook for food safety.2008.New York:Facts on File.
3. Bonardi S, Salmi F, Riboldi E.Detection and count of Salmonella enterica
in pork and poultry meat products.Vet Res Commun2008;32:315–317.
4. Amagliani G, Brandi G, Schiavano GF.Incidence and role of Salmonella
safety.Food Res Int2012;45:780–788.
5. .The community summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in the European Union in 2008.EFSA J2010;8:1496.
6. Iwamoto M, Ayers T, Mahon BE, Swerdlow DL.Epidemiology of seafood
-associated infections in the United States.Clin Microbiol Rev2010;23:399–411.
7. Davies RH, Dalziel R, Gibbens JC.National survey for Salmonella
in pigs, cattle and sheep at slaughter in Great Britain (1999–2000).J Appl Microbiol2004;96:750–760.
8. Feldsine PT, Lienau AH, Leung SC.Detection of Salmonella
in fresh cheese, poultry products, and dried egg products by the ISO 6579 Salmonella
culture procedure and the AOAC official method: collaborative study.J AOAC Int2003;86:275–295.
9. Van der Zee HCorry JEL, Curtis GDW, Baird RM.Media for the isolation of Salmonella spp.
Handbook of culture media for food microbiology2003;372nd ed..Amsterdam:Elsevier;195–208.
10. Kumar R, Surendran PK, Thampuran N.Evaluation of culture, ELISA and PCR assays for the detection of Salmonella
.Lett Appl Microbiol2008;46:221–226.
11. Moreira AN, Conceicao BF, Conceicao R.Monoclonal antibodies against serogroup B Salmonellae:
production, characterisation and use in a sandwich ELISA.Food Agric Immunol2008;19:1–10.
12. Eigner U, Reissbrodt R, Hammann R.Evaluation of a new chromogenic medium for the isolation and presumptive identification of Salmonella
species from stool specimens.Eur J Clin Microbiol Infect Dis2001;20:558–565.
13. Perry JD, Freydiere AM.The application of chromogenic media in clinical microbiology.J Appl Microbiol2007;103:2046–2055.
14. Harrigan WF.Laboratory methods in food microbiology1998.San Diego:Academic Press.
16. Rall R, Mores R, Casale A.Evaluation of three enrichment
broths and five plating media for Salmonella
detection in poultry.Braz J Microbiol2005;36:147–150.
17. Wray C, Wray A.Salmonella
in domestic animals2000.New York:CABI Publishing.
18. Forbes BA, Sahm DF, Weissfeld AS.Bailey and Scott’s diagnostic microbiology2007.St Louis:Mosby.
19. Daniel W.Biostatistics a foundation for analysis in the health science1995.New York:John Wiley and Sons Inc.
20. Heinitz ML, Ruble RD, Wagner DE, Tatini SR.Incidence of Salmonella
in fish and seafood
.J Food Prot2000;63:579–592.
21. Brands DA, Inman AE, Gerba CP, Mare CJ, Billington SJ, Saif LA, et al..Prevalence of Salmonella
spp. in oysters in the United States.Appl Environ Microbiol2005;71:893–897.
22. Kumar R, Surendran PK, Thampuran N.Distribution and genotypic characterization of Salmonella
serovars isolated from tropical seafood
of Cochin, India.J Appl Microbiol2009;106:515–524.
23. .FAO expert workshop on the application of biosecurity measures to control Salmonella c
ontamination in sustainable aquaculture. FAO Fisheries and Aquaculture Report No. 9372010.Mangalore, India:FAO;39.
24. Mansour N, Yassien N, Ibrahim HM, Elmossalami MK.Microbial hazard of salted OM El-Kholoul (wedge shell Donax trunculus
).Zentralbl Veterinarmed B1998;45:53–57.
25. Huss HH.Control of indigenous pathogenic bacteria in seafood
26. Sivapalasingam S, Friedman CR, Cohen L, Tauxe RV.Fresh produce: a growing cause of outbreaks of foodborne illness in the United States, 1973 through 1997.J Food Prot2004;67:2342–53.
27. Smith KM, Sahyoun NR.Fish consumption: recommendations versus advisories, can they be reconciled?Nutr Rev2005;63:39–46.
28. Schrank IS, Mores MAZ, Costa JLA.Influence of enrichment
media and application of a PCR based method to detect Salmonella
in poultry industry products and clinical samples.Vet Microbiol2001;82:45–53.
29. June GA, Sherrod PS, Hammack TS.Relative effectiveness of selenite cystine broth, tetrathionate broth and Rappaport-Vassiliadis medium for recovery of Salmonella
spp from raw flesh highly contaminated food and poultry feed: collaborative study.J AOAC Int1995;79:1307–1323.
30. Rybolt ML, Wills RW, Byrd JA.Comparison of four Salmonella
isolation techniques in four different inoculated matrices.Poult Sci2004;83:1112–1116.
31. Rybolt ML, Wills RW, Bailey RH.Use of secondary enrichment
for isolation of Salmonella
from naturally contaminated environmental samples.Poult Sci2005;84:992–997.
32. Hara-Kudo Y, Nishina T, Nakagawa H.Improved method for detection of Vibrio parahaemolyticus
.Appl Environ Microbiol2001;67:5819–5823.
33. Schönenbrücher V, Mallinson ET, Bülte M.A comparison of standard cultural methods for the detection of foodborne Salmonella
species including three new chromogenic plating media.Int J Food Microbiol2008;123:61–66.
34. Perez JM, Cavalli P, Roure C.Comparison of four chromogenic media and Hektoen agar for detection and presumptive identification of Salmonella
strains in human stools.J Clin Microbiol2003;41:1130–1134.
35. Baraheem OH, El-Shamy H, Bakr W.Bacteriological quality of some dairy products (kariesh cheese and ice cream) in Alexandria.J Egypt Public Health Assoc2007;82:491–510.
36. Maddocks S, Olma T, Chen S.Comparison of CHROMagar
Salmonella medium and xylose-lysine-desoxycholate and Salmonella–Shigella
agars for the isolation of Salmonella
strains from stool samples.J Clin Microbiol2002;40:2999–3003.
37. Gaillot O, Di Camillo P, Berche P.Comparison of CHROMagar
Salmonella medium and Hektoen enteric agar for isolation of Salmonellae
from stool samples.J Clin Microbiol1999;37:762–765.