Characterization of coagulase-positive staphylococci isolated from tank and silo ewe milk.

The prevalence of coagulase-positive staphylococci (CPS) was studied among 390 samples of ewe milk. Fifty-seven (14.85%) samples of tank milk and all samples (6) of silo milk gave a positive result on Baird-Parker agar base supplemented with rabbit plasma fibrinogen, whereas amplification of the coagulase (coa) gene was successful in 6 (1.56%) samples of tank milk and in all silo samples. Phenotypic and genetic analysis of 153 isolates identified 151 (98.69%) as Staphylococcus aureus. Amplification of the coa gene was positive for 149 isolates (97.39%). The staphylococcal enterotoxin (SE) C gene was detected in 116 strains (75.82%), whereas more than one SE gene was carried in 5 strains (3.26%). None of the isolates harbored the genes for SEE or for methicillin resistance. A high prevalence of CPS carrying enterotoxin genes should be of concern because ewe milk is mainly processed into raw milk cheeses. The detection of the coa gene from milk samples could help to assess the microbiological safety of raw milk intended for direct use in the dairy industry.

Identification and epidemiological relationships of Aeromonas isolates from patients with diarrhea, drinking water and foods.

A collection of Aeromonas isolates obtained over a three-year period in the same geographic area (León, NW of Spain) was characterized by (GTG)5-PCR fingerprinting, amplified fragment length polymorphism (AFLP) analysis and gyrB gene sequence analysis. The isolates originated from human diarrheal stools (29 isolates), potable water (13 isolates), rabbit meat (13 isolates) and marine fish (5 isolates). The distribution of Aeromonas species varied with the strain source. Aeromonas caviae HG4 and Aeromonas media HG5 were predominant in clinical and water isolates, respectively, whereas motile Aeromonas salmonicida HG3 strains were most frequently found in fish and meat. Molecular typing revealed several genotypic relationships among specific isolate subsets: (i) two clones of A. media HG5 persisted in drinking water over the study period, (ii) different patients harbored identical or closely related clones during several months, and (iii) clonal relatedness was observed in two sets of water and human isolates. The first of these sets comprised nine water isolates and two human A. media HG5 isolates, whereas the other one included a water isolate and a human isolate of A. caviae HG4. The latter finding suggests that Aeromonas transmission in the studied region followed a waterborne route. Interestingly, the three human isolates closely related to water isolates were recovered in a period of four days in June 2006 from non-related patients without underlying medical conditions that tested negative for other enteric pathogens. The data imply the transmission through contaminated water of strains of the A. caviae group that can produce disease in humans.

Identity, virulence genes, and clonal relatedness of Aeromonas isolates from patients with diarrhea and drinking water.

Among 800 stool specimens from patients with diarrhea submitted by Primary Care Centers for routine analysis to the Hospital of León (NW Spain) Microbiology and Parasitology Service, 32 (4%) were tested positive for Aeromonas spp. Mixed infections with other enteric pathogens occurred in 12 patients. A. caviae was isolated from 23 clinical specimens. There were also patients infected with A. media, A. hydrophila, A. bestiarum, and A. veronii biovar veronii. All but three isolates carried one or more of the virulence genes. The incidence of the alt, hlyA, aerA, ast, and laf genes was 71.9, 28.1, 25.0, 18.8, and 9.4%, respectively. The alt(+)/ast(+) combination was detected in four isolates and the aerA(+)/hlyA(+) combination was detected in the two A. hydrophila isolates. None of the strains harbored the TTSS, stx1, or stx2 genes and nine bore plasmids. Thirty clinical isolates and a collection of 12 A. caviae and A. media strains obtained from León municipal drinking water over the study period were typed by pulsed-field gel electrophoresis (PFGE). PFGE patterns revealed genetic relatedness and persistence over time among water isolates and some clinical isolates. Interestingly, one A. caviae (aerA(-)/hlyA(-)/alt(+)/ast(-)/laf(+)) human isolate and two A. caviae (aerA(-)/hlyA(-)/alt(+)/ast(-)/laf(+)) drinking water isolates had indistinguishable PFGE patterns, suggesting waterborne infection.

Foodborne and indicator bacteria in farmed molluscan shellfish before and after depuration.

Galicia's coast (northwestern Spain) is a major producer of bivalve molluscs. Over an 18-month period, the presence of Salmonella, Aeromonas, Plesiomonas shigelloides, Vibrio parahaemolyticus, and Clostridium botulinum was determined by PCR methods in mussels (22 batches) and infaunal bivalves (31 batches of clams and cockles) before and after depuration. All batches were harvested from Galician class B harvesting areas where bivalve molluscs must not exceed 4,600 Escherichia coli per 100 g of flesh and liquor in 90% of the samples. Virulence-associated genes of Salmonella (invA), Aeromonas (aerA, hlyA, alt, ast, and laf), P. shigelloides (hugA), V. parahaemolyticus (tdh and trh), and C. botulinum (BoNT) were not detected. The pR72H chromosomal DNA fragment, which is conservative in V. parahaemolyticus strains, was detected in five (4.7%) samples. A number of 192 suspect isolates did not fit the description of clinical Aeromonas phenospecies, pathogenic Vibrio spp., or P. shigelloides. The effectiveness of commercial depuration in reducing bacterial indicators was also examined. E. coli was reduced to < or = 230/100 g of flesh and liquor in 90.9% of mussel lots but in only 70.9% of infaunal bivalve lots. For total coliform elimination, mussels were also more effective. Total counts significantly (P < 0.005) correlated with numbers of Pseudomonas, Aeromonas, and Vibrio. Our data indicate that Salmonella and pathogenic bacteria indigenous to estuarine environments do not appear to be significant hazards in Galician molluscan shellfish. A reason for concern, however, is that clearance of E. coli to acceptable levels was not always achieved especially in infaunal bivalves.

Occurrence of foodborne pathogenic bacteria in retail prepackaged portions of marine fish in Spain.

AIMS: To survey the presence of indigenous and nonindigenous foodborne bacterial pathogens in displayed prepacked portions of fresh marine fish. METHODS AND RESULTS: A survey of 50 different samples of fresh marine fish (conger, swordfish, sole, grouper and whiting) was conducted over a period of 5 months. Trays of fillets and steaks were obtained at retail level and tested for foodborne bacterial pathogens. Vibrio cholerae and Salmonella were not detected. Two samples (4%) yielded Vibrio strains carrying a DNA fragment specific for Vibrio parahaemolyticus, but resulted negative to PCR amplification of the virulence-related tdh gene. Levels of motile Aeromonas ranging from 2.29 to 7.20 log CFU g(-1) were found in 31 (62%) samples. All fish portions were positive for the Aeromonas hlyA gene and 38 for both aerA and hlyA genes, which may contribute to diarrhoea-related virulence. The incidence of Listeria monocytogenes was 10%. Levels of Staphylococcus aureus lower than 2 log CFU g(-1) were found in 15 (30%) samples. Numbers of presumptive Clostridium perfringens ranging from 1.82 +/- 0.22 to 4.26 +/- 1.25 log CFU g(-1) were detected in 42 (84%) samples. Edwardsiella tarda was detected in two samples of grouper fillets. CONCLUSIONS: Displayed portions of raw fish carried bacteria that can cause foodborne disease. The risk posed by fresh fish when properly cooked is low, but high when destined to be consumed raw, undercooked or very lightly processed. SIGNIFICANCE AND IMPACT OF THE STUDY: This study revealed that raw fish sold in Spain could be a source of foodborne bacterial pathogens. Improvements in handling and processing are needed to minimize the prevalence of pathogenic bacteria.

Molecular and phenotypic characterization of nonmotile Gram-negative bacteria associated with spoilage of freshwater fish.

AIMS: Characterization of nonmotile bacteria associated with freshwater fish spoilage and that phenotypically resembled Psychrobacter spp. METHODS AND RESULTS: A population of 44 nonmotile Gram-negative rods could not be assigned to the genus Psychrobacter on the basis of a definitive test (transformation assay). Conventional and commercial phenotypic systems did not help in identification. A second extensive phenotypic analysis using different temperatures and media confirmed these isolates as nonmotile although electron microscopic examination showed that all but two had one to four polar flagella and other appendages. On the basis of numerical taxonomy, this population was divided into six clusters, one of them consisting of five fluorescent strains. Sequencing of fluorescent and non fluorescent representative strains from each cluster demonstrated that strains from five clusters had between 97.8 and 98.8% sequence homology with Pseudomonas fragi IFO 3458. This and an unknown strain from deep sea were the closest organisms (80.9% sequence homology) to one aflagellated representative strain of the remaining cluster. CONCLUSIONS: Oxidase-positive, nonmotile, nonfermenter Gram-negative rods isolated from freshwater fish can be wrongly ascribed to the genus Psychrobacter. SIGNIFICANCE AND IMPACT OF THE STUDY: Molecular methods are necessary for the identification of environmental isolates and species with an incomplete phenotypic description. This work emphasizes the need for a sound description of Ps. fragi based on molecular and phenotypic characterization.

PCR detection of potentially pathogenic aeromonads in raw and cold-smoked freshwater fish.

AIMS: Development of a PCR assay for detection of aeromonads carrying the hlyA and/or aerA genes in fish. METHODS AND RESULTS: The protocol involves an overnight selective enrichment step in tryptic soy broth yeast extract containing 10 microg ml(-1) of ampicillin followed by extraction of DNA and PCR amplification of two haemolysin genes that contribute to the virulence of Aer. hydrophila. This procedure can detect initial populations of 1-10 cfu g(-1) within 24 h in artificially contaminated samples. In naturally contaminated fish, both genes were detected in 13 out of 14 fresh fish lots (aeromonads levels between < 1 and 5.42 log cfu g(-1)) and in 4 out of 16 lots of vacuum-packed cold-smoked fish (aeromonads levels between < 1 and 3.37 log cfu g(-1)). Before enrichment, dominant species were Aer. hydrophila HG1 (aerA+hlyA+), Aer. bestiarum HG2 (aerA+hlyA+) and Aer. caviae HG4 (aerA-hlyA-). After enrichment, Aer. hydrophila HG1 (aerA+hlyA+) was dominant. CONCLUSIONS: Fresh fish and even smoked fish carry hlyA+ and/or aerA+ aeromonads that can be detected by PCR within 24 h. SIGNIFICANCE AND IMPACT OF THE STUDY: The PCR assay described offers considerable potential as a rapid method with specificity, sensitivity and simplicity.

Evaluation of the spiral plating system for the routine assessment of indicator microorganisms in raw ewe's milk.

The numbers of members of different microbial groups in bulk raw ewe's milk used for cheesemaking (46 samples, taken on receipt at the dairy over 1 year) were assayed by the spiral plating system to determine the effectiveness of this method compared with that of widely accepted conventional methods in providing counts. The results indicated that for ewe's milk, the suitability of the spiral plating system depends to a great extent on the microbial group studied. Although "spiral" counts of mesophiles, psychrotrophs, coliforms, and Enterobacteriaceae could be considered equivalent to those obtained by conventional techniques (r > or = 0.90; variance between replicate platings approximately 0.005), the automated method was found not to be suitable for the assessment of other groups of indicator bacteria (thermodurics and enterococci). Counts of lactic acid bacteria and yeasts and molds were affected significantly (P < 0.05) by the plating method, although other statistical parameters were more favorable (r = 0.88 and r = 0.82, respectively; 95% confidence limits within 0.5 log units). Finally, counts of staphylococci, particularly on Baird-Parker medium, showed less variation and higher reproducibility with the spiral method. Nevertheless, for the routine microbiological analysis of ewe's milk, the spiral plating system, with its time-, effort-, and material-saving advantages, is preferred over the conventional method.

Virulence markers in Aeromonas hydrophila and Aeromonas veronii biovar sobria isolates from freshwater fish and from a diarrhoea case.

AIMS: To evaluate the public health significance of representative strains of two Aeromonas spp., mainly from freshwater fish, on the basis of production of virulence-associated factors and presence of the haemolytic genes aerA and hlyA. METHODS AND RESULTS: Eleven strains of Aer. hydrophila, three strains of Aer. veronii biovar sobria (all from freshwater fish) and one strain of Aer. hydrophila from human diarrhoea were tested for potential virulence traits and for the presence of the haemolytic genes aerA and hlyA. Ten Aer. hydrophila isolates were aerA(+)hlyA(+) and two aerA(+)hlyA(-). Aeromonas veronii biovar sobria isolates were aerA(-)hlyA(-). Strains from the three genotypes showed enterotoxic activity in the suckling mouse assay. At 28 degrees C, four Aer. hydrophila fish strains could be considered as potentially virulent (possessing at least two of these characteristics: haemolytic, cytotoxic and enterotoxic). One Aer. veronii biovar sobria strain and the clinical isolate were cytotoxic on Vero cells. When grown at 4 degrees C, these six isolates fulfilled virulence criterion, but at 37 degrees C, only one fish strain, an Aer. hydrophila, did. CONCLUSIONS: The potential health risk derived from the presence of Aer. hydrophila and Aer. veronii biovar sobria in ice-stored freshwater fish should not be underestimated. SIGNIFICANCE AND IMPACT OF THE STUDY: Expression of virulence factors is affected by temperature incubation and not always related to the presence of haemolytic genes.

Foodborne pathogenic bacteria in prepackaged fresh retail portions of farmed rainbow trout and salmon stored at 3 degrees C.

Twelve lots of fresh unskinned fillets of rainbow trout (Oncorhynchus mykiss) and 10 lots of fresh sliced salmon (Salmo salar) prepacked in trays wrapped with an oxygen-permeable film were obtained immediately after packing from two supermarkets having in-plant facilities for packaging wet fish. During storage at 3 degrees C, Listeria innocua was detected in eight lots of trout fillets after 4 days storage. L. monocytogenes was recovered from a single lot also contaminated with L. innocua. Initial numbers of aeromonads were significantly (p < 0.05) lower in trout fillets (3.35 +/- 0.62 log cfu g(-1)) than in salmon slices (4.20 +/- 0.89 log cfu g(-1)). In both fish products, these bacteria significantly (p < 0.05) increased up until spoilage. Most Aeromonas spp. isolates from trout fillets were assigned to A. veronii biovar sobria HG8 (hybridisation group 8), A. caviae HG4, A. eucrenophila HG6, A. hydrophila HG1 and A. veronii biovar veronii HG10. Strains of HG12 (A. schubertii), HG4 and HG8 formed the majority of aeromonads recovered from salmon slices.

Comparison of the acidifying activity of Lactococcus lactis subsp. lactis strains isolated from goat's milk and Valdeteja cheese.

AIMS: This work was carried out to study the acid production by Lactococcus lactis subsp. lactis strains isolated from goat's milk and goat cheese (Valdeteja variety) in order to select a suitable starter culture for industrial goat cheese manufacturing. METHODS AND RESULTS: The titrable acidity of 45 Lactococcus lactis subsp. lactis strains isolated from a home-made batch of Valdeteja cheese with excellent sensory characteristics was measured over a period of 18 h. The strains were divided into two groups depending on the acid production rate: 20 fast acid producer (F) strains and 25 slow acid producer (S) strains. The kinetic parameters (lag phase, maximum acid production rate and value of upper asymptote curve) of the acid production curves for F and S strains were significantly (P < 0.001) different. CONCLUSIONS: Significant (P < 0.001) differences between titrable acidity of F and S strains were observed after the second hour of incubation. SIGNIFICANCE AND IMPACT OF THE STUDY: An F strain acetoin producer (Lactococcus lactis subsp. lactis 470Ch2) was selected as autochthonous starter culture for industrial Valdeteja goat cheese manufacturing.

Surface mycoflora of a Spanish fermented meat sausage and toxigenicity of Penicillium isolates.

The surface mycoflora of "chorizo de Cantimpalos", a Spanish variety of fermented meat sausage characterised by a natural white covering, has been investigated. Among 54 mould strains isolated, 38 belonged to Penicillium subgenus Penicillium. The major species found (18 isolates) was identified as Penicillium commune, and the other dominant species (13 isolates) was identified as P. olsonii. None of the P. olsonii isolates produced cyclopiazonic acid, mycophenolic acid, roquefortine C. patulin or ochratoxin A, but all P. commune isolates produced cyclopiazonic acid. Toxicity to Artemia salina larvae was very high for all P. commune isolates investigated, while no isolates of P. olsonii studied were toxic to these crustaceans. The results may assist in selection of nontoxic strains, which could be used as surface starters in the manufacture of this type of sausage. The apparent inability to produce penicillin is a valuable characteristic to take into account in the selection process.

Bacteriological quality of aquacultured freshwater fish portions in prepackaged trays stored at 3 degrees C.

Fresh trout fillets and salmon slices packed in trays were obtained from two multinational chain supermarkets and evaluated for freshness and bacteriological quality immediately after packaging and during storage at 3 degrees C. Initial aerobic counts at 30 and 25 degrees C were significantly (P < 0.05) lower in trout fillets (5.27 +/- 0.57 and 4.87 +/- 0.80 log CFU/g, respectively) than in salmon slices, where levels in excess of 6 log CFU/g were found. In both products, initial Enterobacteriaceae counts were slightly higher than 3 log CFU/g and increased significantly during shelf life by approximately 3 log CFU/g. Most of the enterobacteria were identified as Citrobacter freundii, Hafnia alvei, and Enterobacter cloacae. On day 0, most probable number (MPN) counts of total and fecal coliforms were not significantly different, numbers of the latter group being approximately 4 MPN/g. Escherichia coli was only detected when fish was spoiled. Although initial presumptive Staphylococcus aureus counts were approximately 3 log CFU/g, only 4 of 84 selected colonies belonged to this species. Neither Salmonella nor antimicrobial residues were detected in any sample. Ethanol content in salmon slices did not significantly (P > 0.05) increase until they became inedible. Significant correlation (r = +0.72, P < 0.05) was observed between this chemical index and viable counts at 30 degrees C only when salmon slices were inedible. Trout fillets were acceptable for 7 days, and salmon slices showed signs of spoilage after 4 days. Although public health concerns associated with packed trout and salmon appear to be minimal, data on sensory quality, shelf life, and total viable and Enterobacteriaceae counts strongly suggest the need to improve the quality control systems used by European multinational retailers, especially for imported salmon.

Mesophilic aeromonads in wild and aquacultured freshwater fish.

Numbers and species of motile Aeromonas were determined in freshly caught freshwater fish, in the surrounding environment, and also during iced chilled storage of fish specimens. Although no significant differences were observed in water samples, initial levels for skin, gill, and intestines were significantly lower in farmed rainbow trout (Oncorhynchus mykiss) than in wild brown trout (Salmo trutta) and pike (Esox lucius). During storage of wild specimens, naturally occurring aeromonads grew fairly well on the surfaces of skin and body cavity. Of 171 strains assigned to the genus Aeromonas, 88% were identified to phenospecies and putative genospecies level by using comprehensive biochemical schemes. The isolates were allocated to putative hybridization groups (HGs) 1 and 3 Aeromonas hydrophila (29%); putative HG 8 Aeromonas veronii biovar sobria (19%); putative HG 2 Aeromonas bestiarum (18%); putative HG 9 Aeromonas jandaei (16%); putative HGs 4 and 5a Aeromonas caviae (2%); putative HG 12 Aeromonas schubertii (2%); and putative HG 11 (unnamed, 0.6%). The remaining 20 isolates (12%) resembled A. schubertii but could not be allocated to currently recognized phenospecies or to putative HGs. Although cultured rainbow trout yielded strains of putative HGs 1, 4, and 8, which appear to be of major clinical importance, most isolates assigned to putative HGs 1 and 8 were recovered from pike. Differences among HGs found in wild animals could be related to their origin (unpolluted rivers for brown trout and urban rivers for pike). The recovery of these aeromonads species was not related to sampling site. The initial levels of motile aeromonads, their behavior during storage, and the strong potential spoilage activity of most isolates confirm that these bacteria can contribute to deterioration of iced wild freshwater fish. Although adequate cooking would inactivate motile aeromonads, the high incidence of isolates belonging to gastroenteritis-associated HGs should be regarded as a potential health concern, particularly for susceptible populations when there is a possibility of cross-contamination.

Psychrobacters and related bacteria in freshwater fish.

Three phenotypic identification systems were employed to identify 106 strains of gram-negative, nonmotile, aerobic bacteria obtained during iced storage of wild (Salmo trutta and Esox lucius) and farmed (Oncorhynchus mykiss) freshwater fish. Using diagnostic tables and computer-assisted identification, the isolates were Psychrobacter (64 strains), Acinetobacter (24 strains), Moraxella (6 strains), Chryseobacterium (5 strains), Myroides odoratus (2 strains), Flavobacterium (1 strain), Empedobacter (1 strain), and unidentified (3 strains). Overall similarities of all strains were determined for 108 characters by numerical analysis (simple matching coefficient of similarity [S] and clustering by unweighted pair group average linkage [UPGMA]). At the 77% similarity level, 92 strains formed nine major clusters (3 or more strains) and four small clusters (2 strains). Cluster 1 (25 isolates divided into two main subclusters) could be assigned to Psychrobacter phenylpyruvicus, clusters 2 and 3 (26 isolates) were designated as Psychrobacter immobilis, and clusters 4 (3 isolates) and 7 (4 isolates) were identified as Psychrobacter urativorans and Psychrobacter spp., respectively. Clusters 5 (five isolates), 6 (three isolates), and 9 (five isolates) were labeled as Acinetobacter spp., Acinetobacter johnsonii, and Acinetobacter lwoffii, respectively. Cluster 8 (12 isolates), with a high resemblance to Thornley's phenon 4 (a heterogeneous group of bacteria isolated from poultry and related to Acinetobacter), remained unnamed. The restriction pattern was identical for strains grouped into clusters 2 and 3 (P. immobilis) but was different for the remaining Psychrobacter isolates. A large proportion of isolates belonging to the family Moraxellaceae were closely related. Psychrobacters and A. johnsonii were present in freshly caught fish and river water. In the latter stages of storage, P. phenylpyruvicus and acinetobacters tended to decrease, whereas P. immobilis increased.

Yeast populations on Spanish fermented sausages.

Yeast populations on 24 lots of Spanish fermented sausages, made by four factories (F1, F2 and F4, artisanal; F3, industrial) were investigated throughout manufacture and the influence of different variables evaluated. In addition, 41 yeast strains were identified at the species level using two miniaturised systems: ATB32C (API System) and Vitek Yeast Biochemical Card (Vitek YBC). Levels of yeasts found in the sausage mixture (mean counts around 4 log units/g) were similar to those described by other authors. In sausages from factories F1 and F2, a further increase was noted, reaching 5.5 log units/g after fermentation. Counts subsequently decreased to 3.6 and 5 log units/g, respectively. In sausages from factories F3 and F4, decreasing counts were observed from the beginning, particularly in sausages from F3, where yeasts were almost absent in the finished product. Type of manufacture and sausage diameter, were the variables most influencing yeast counts. Debaryomyces hansenii (teleomorph of C. famata) was the dominant species, being found at all stages of manufacture. Trichosporon ovoides (formerly T. beigelii), Yarrowia lipolytica (perfect form of C. lipolytica), C. intermedia/curvata, C. parapsilosis, C. zeylanoides and Citeromyces matritensis (teleomorph of C. globosa) were also present. Direct identification was possible only with 50% of the total of strains investigated, although a higher number of strains was identified using the API than the Vitek YBC system.

Hemolytic and elastolytic activities influenced by iron in Plesiomonas shigelloides.

The aim of this study was to determine the presence of hemolytic and elastolytic enzymes in several strains of Plesiomonas shigelloides in relation to the availability of iron in culture media. Hemolytic activity and elastolytic activity were detected in strains of P. shigelloides and were enhanced when the strains were grown in an iron-depleted medium and lost after thermal treatment at 100 degrees C for 10 min. Also, elastolytic activity was inactivated by phenylmethyl sulfonyl fluoride, an inhibitor of serine proteases. Hemolytic activity was detected extracellularly in cell-free supernatants, whereas elastin degradation activity was cell associated. Both activities may be related to the virulence of P. shigelloides.

Hemolytic activity and siderophore production in different Aeromonas species isolated from fish.

The hemolytic activity and siderophore production of several strains of motile aeromonads were determined. The hemolytic activity of Aeromonas caviae and Aeromonas eucrenophila was enhanced after trypsinization of the samples. The enhancement of hemolysis was observed in strains that carried an aerolysin-like gene, detected by a PCR procedure. Siderophore production was demonstrated in all but one strain of Aeromonas jandaei. No apparent relationship was observed between the presence of plasmid DNA and hemolysis or siderophore production.

Bacterial microflora of wild brown trout (Salmo trutta), wild pike (Esox lucius), and aquacultured rainbow trout (Oncorhynchus mykiss).

Initial numbers of bacteria associated with wild (brown trout and pike) and cultured (rainbow trout) freshwater fish as well as with the water in which they were caught were determined. Subsequently, a total of 979 randomly selected isolates were characterized and identified to the genus level. For all counts performed (aerobes, psychrotrophs, anaerobes, Enterobacteriaceae, and enterococci), no significant differences were observed in water samples, the highest level corresponding to psychrotrophs in pike environments (4.23 X 10(3) CFU/ml). Overall, the skin and intestinal content of brown trout were the most contaminated, while rainbow trout specimens (gills and gut) yielded the lowest numbers. For all bacterial groups, pike gills had the highest numbers. Counts for all of the sampling sites compare well with findings in other temperate geographical environments. Biological characteristics (feeding and skin properties) and the use of antimicrobials in aquaculture might have influenced these results. Motile and nonmotile aerobic gram-negative bacteria together with Enterobacteriaceae accounted for 50 to 70% of the psychrotrophs isolated from water. Micrococcaceae, lactic acid bacteria, Bacillus, and coryneforms were also found. The groups represented in psychrotrophic isolates from the outer surfaces do not reflected those detected in water, so it was common that those organisms recovered in significant numbers from fish were not detected in surrounding habitat of the fish. Motile aeromonads and Carnobacterium were the dominant psychrotrophs in the guts of pike and brown trout, respectively. The intestinal content of reared fish gave a high incidence of Bacillus and coryneforms, while Enterobacteriaceae was absent. Again, rearing practices could have influenced this finding. Listeria monocytogenes was not detected in any of the examined samples. Two strains of Salmonella, which belonged to the same serovar and lysotype, were recovered from pond-water samples taken from one facility on different sampling days. From the gut of a pike specimen and from the pike's environment, two Plesiomonas shigelloides strains of different serovars were recovered. These latter four strains were resistant to a considerable number of antimicrobial compounds (multiple antibiotic resistance indices > 0.2).

Numbers and species of motile aeromonads during the manufacture of naturally contaminated Spanish fermented sausages (longaniza and chorizo).

The prevalence of aeromonads in the initial mixes of 14 batches of chorizo and longaniza obtained from three small- and middle-sized factories was 78.5%, with counts ranging from >1.00 to 4.47 log10 CFU/g. Only 2 of 10 mixture samples prepared at a large and modern processing plant yielded these organisms, with levels below 1 log10 CFU/g. The hygienic status of factories significantly affected incidence and counts. Of 39 presumptive isolates from glutamate starch penicillin agar, 36 were confirmed as motile aeromonads and allocated to Aeromonas hydrophila (n = 24), A. veronii biovar sobria (n = 10), and A. caviae (n = 2). All of them were beta-hemolytic and capable of growing at 5 degrees C. Regardless of initial contamination, aeromonads were rapidly inactivated during the early stages of manufacture.