Shelf life and safety aspects of chilled cooked and peeled shrimps (Pandalus borealis) in modified atmosphere packaging
- Department of Seafood Research, Danish Institute for Fisheries Research (DIFRES), Lyngby
- Royal Greenland Seafood Ltd, Glyngøre, Denmark
- Department of Seafood Research, Danish Institute for Fisheries Research (DIFRES), Lyngby
Ole Mejlholm, Danish Institute of Fisheries Research, Department of Seafood Research, c/o Technical University of Denmark, Søltofts Plads, Building 221, DK-2800, Kgs. Lyngby, Denmark (e-mail: email@example.com).
Aims: To evaluate the growth of Listeria monocytogenes and shelf life of cooked and peeled shrimps in modified atmosphere packaging (MAP).
Methods and Results: Storage trials with naturally contaminated cooked and peeled MAP shrimps (Pandalus borealis) were carried out at 2, 5 and 8°C. Challenge tests at the same conditions were performed after inoculation with Listeria monocytogenes. Both storage trials and challenge tests were repeated after 4 months of frozen storage (−22°C). Brochothrix thermosphacta and Carnobacterium maltaromaticum were responsible for sensory spoilage of cooked and peeled MAP shrimps. In challenge tests, growth of L. monocytogenes was observed at all of the storage temperatures studied. At 5 and 8°C the concentration of L. monocytogenes increased more than a 1000-fold before the product became sensory spoiled whereas this was not observed at 2°C. Frozen storage had only a minor inhibiting effect on growth of L. monocytogenes in the thawed product.
Conclusions: To prevent L. monocytogenes becoming a safety problem, cooked and peeled MAP shrimps should be distributed at 2°C and with a maximum shelf life of 20–21 d. At higher temperatures shelf life is significantly reduced.
Significance and Impact of the Study: Information is provided to establish shelf life of cooked and peeled MAP shrimps.
In recent years, interest in mildly preserved convenience foods has increased, and establishment of storage conditions to assure product safety and sufficient shelf life for distribution is an important issue (Gould 2000). Raw shrimps have a very short shelf life and freezing or the combined use of brining and chilling are common preservation methods for shelf life extension of shrimp products (Dalgaard and Jørgensen 2000). Modified atmosphere packaging (MAP) facilitates distribution of seafood from chill cabinets in supermarkets and it has been studied for raw and lightly preserved shrimp products (Dalgaard and Jørgensen 2000; López-Caballero et al. 2002). However, for cooked and peeled MAP shrimps we found no previous studies on the chilled shelf life, growth of spoilage micro-organisms or the important food-borne pathogen Listeria monocytogenes. This ready-to-eat (RTE) product is mildly preserved and frozen; retail distribution prior to chilled distribution in supermarkets could be appropriate to overcome potential shelf life problems. In addition to reducing the time of chilled distribution, frozen storage has previously been reported to inhibit or inactivate both spoilage and pathogen micro-organisms (Guldager et al. 1998; Lund 2000).
Shelf life of chilled raw shrimps is short (6 d at 5–6°C) as their high pH (c. 7·5) allows Gram-negative micro-organisms, including Pseudomonas fragi and Shewanella putrefaciens, to grow rapidly under aerobic storage conditions (Lannelongue et al. 1982; Matches 1982; Chinivasagam et al. 1996). Packaging of raw shrimps in modified atmospheres with elevated concentrations of CO2 extends shelf life and changes the spoilage microflora to become dominated by Gram-positive bacteria or by the CO2-resistant Gram-negative bacterium Photobacterium phosphoreum (Lannelongue et al. 1982; López-Caballero et al. 2002). When cooked, the concentration of micro-organisms on shrimps is markedly reduced but during chilling and peeling, the product can be recontaminated by a predominantly Gram-positive microflora (Harrison and Lee 1968). The effect of MAP on the chilled shelf life and growth of spoilage and potential pathogenic micro-organisms in cooked and peeled shrimps is of practical importance but as indicated above remains to be documented.
Listeria monocytogenes and Clostridium botulinum type E are pathogenic and psychrotolerant bacteria of potential importance in a chilled RTE product such as cooked and peeled MAP shrimps. Listeria monocytogenes can grow close to 0°C and Cl. botulinum type E produce toxin above 3°C (Huss et al. 2004). Listeria monocytogenes was detected in 2% of 3331 samples of cooked and peeled shrimps from Icelandic factories (Valdimarsson et al. 1998). Clearly, it is important to limit growth of L. monocytogenes in this product as high concentrations in foods may cause serious illness with fatality rates of 20–40% for susceptible consumers (Farber and Peterkin 2000). Clostridium botulinum type E is indigenous to the aquatic environment where coldwater shrimps are caught and the boiling procedure used for shrimps is insufficient to inactivate the spores of Cl. botulinum type E (Huss et al. 2004). Consequently, an evaluation of the growth of L. monocytogenes and toxin formation by Cl. botulinum type E in chilled cooked and peeled MAP shrimps is important for safety of this product.
The objectives of the present study were to evaluate shelf life and safety of cooked and peeled MAP shrimps. The product was studied by storage trials and challenge tests carried out just after packing in a modified atmosphere and again after 4 months frozen storage of the MAP product. First, shelf life and microbiological and chemical changes were determined in storage trials at 2, 5 and 8°C. Secondly, growth of L. monocytogenes was evaluated in challenge tests at 2, 5 and 8°C. Finally, growth of L. monocytogenes observed in the present study was compared with available predictive models, and toxin formation by Cl. botulinum was predicted for the product characteristics and storage conditions of the cooked and peeled MAP shrimps.
Materials and methods
Two series of storage trials with cooked and peeled shrimps packed in a modified atmosphere were carried out during April–May 2003, without frozen storage, and during August–September 2003 after approx. 4 months frozen storage of the MAP product. For both series of experiments individually quick frozen shrimps (Pandalus borealis) from the North Atlantic Ocean were used. The shrimps were catched, cooked, peeled, frozen and glazed in Greenland, and supplied in the frozen state to Danish Institute for Fisheries Research (DIFRES) by a Danish shrimp processor. At DIFRES the shrimps were kept at −22°C until the start of the experiments. Shrimps were randomly divided into six sub-batches and approx. 115 g of frozen shrimps were packed in a modified atmosphere initially containing 50% CO2, 30% N2 and 20% O2 (AGA Ltd, Copenhagen, Denmark). A Multivac C500 packaging machine (Multivac, Vejle, Denmark) and packaging film (NEN 40 HOB/LLPDE 75, Amcore Flexibles, Horsens, Denmark) with low gas permeability (0·45 ± 0·15 cm3 m−2 for O2 and 1·8 ±0·6 cm3 m−2 for CO2) were used for packaging. The gas/shrimp ratio was 4/1 in the final packed product. During handling and packaging the shrimps remained in the frozen state. Immediately after packaging three sub-batches were stored at 2, 5 and 8°C, respectively, whereas the remaining three sub-batches were kept frozen for 4 months at approx. −22°C before chilled storage at 2, 5 and 8°C. For all sub-batches the temperature of the shrimps was recorded continuously throughout frozen and/or chilled storage by data loggers (Tinytag, Gemini Data Loggers Ltd, Chichester, UK). At regularly intervals during the chilled storage three packs from each sub-batch were analysed by microbiological and chemical methods whereas two packs were analysed by sensory evaluation.
Microbiological analyses. Twenty grams of shrimp were diluted 10-fold in chilled (5°C) physiological saline (0·85% NaCl) with 0·1% peptone (PS) and homogenized 60 s in a Stomacher 400 (Seward Medical, London, UK). Further appropriate 10-fold dilutions of the homogenates were made in chilled PS. Aerobic Plate Counts (APC) were determined by spread plating (15°C, 7 d) on Long and Hammer agar (LH) with 1% NaCl (van Spreekens 1974) and lactic acid bacteria (LAB) were enumerated by pour plating (25°C, 3 d) in nitrite actidione polymyxin agar (NAP) with pH 6·2 (Davidson and Cronin 1973). At the time of sensory rejection Lactobacillus spp. were determined by spread plating (25°C, 5 d) on Rogosa agar (Oxoid CM627) incubated anaerobically (5–10% CO2/90–95% N2); Enterococcus were determined by spread plating (44–45°C, 2 d) on Slanetz and Bartley agar (CM0377); Enterobactericeae were enumerated by spread plating (25°C, 2 d) on tryptone soya agar (Oxoid CM0131) covered with a double layer of violet red bile glucose (VRBG) agar (Oxoid CM0485); H2S-producing micro-organisms were determined by pour plating (25°C, 3 d) in iron agar Lyngby (IA) (Oxoid CM964); Brochothrix thermosphacta were enumerated by spread plating (25°C, 2–3 d) on streptomycin-sulphate, thallous-acetate and actidione (STAA) agar (Oxoid CM881) with STAA Selective Supplement (Oxoid SR0151) and L. monocytogenes were determined qualitatively in 25 g of sample as described previously (Jørgensen and Huss 1998). To characterize the spoilage microflora, c. 20 colonies from each sub-batch were isolated from LH at the time of sensory rejection of the shrimps. All types of colonies growing on the plates were represented among the isolates, in numbers reflecting their relative proportion of the total microflora. The isolates were initially characterized by: colony morphology, cellular morphology and motility by the use of phase-contrast microscopy, Gram reaction (KOH method), catalase (H2O2 method), cytochrome oxidase (DryslideTM Oxidase strips, Difco Laboratories), metabolism of glucose (Hugh and Leifson 1953) and growth on STAA agar. Based on these data the isolates were divided into: (i) LAB being Gram-positive, fermentative, catalase and cytochrome oxidase-negative rods and cocci; (ii) Brochothrix-like being Gram-positive, fermentative, positive for growth on STAA agar, catalase-positive and cytochrome oxidase-negative rods; and (iii) Gram-negative isolates being catalase and cytochrome oxidase positive. LAB isolates were further characterized by growth on acetate agar, final pH in La-broth, production of NH3 from arginine (aerobically and anaerobically with 0·1 and 2·0% glucose), production of gas from glucose and gluconate, production of acetoin (Voges-Proskauer) and fermentation of inulin, lactose, mannitol, methyl-α-d-glucoside, methyl-α-d-mannoside and xylose. Tests were carried out as previously described (Wilkinson and Jones 1977; Dalgaard et al. 2003) and isolates identified according to Collins et al. (1987) and Mora et al. (2003). Reference and type strains of Carnobacterium piscicola (DSM 20730T), C. divergens (DSM 20623T), Lactobacillus curvatus subsp. curvatus (DSM 20019T) and Lact. sakei subsp. sakei (DSM 20017T) were included for comparison. Brochothrix-like isolates were further characterized by growth with 1·0, 6·5 and 8·0% NaCl, growth with 0·5% potassium telluride and fermentation of rhamnose. Brochothrix-like isolates were identified according to Wilkinson and Jones (1977) and Holley (2000). The type strain of B. thermosphacta (ATCC 11509T) was included. Gram-negative isolates were further identified for vibriostaticum sensitivity (150 μg), production of indole, reduction of trimethylamine-oxide (TMAO) and production of H2S. In addition, for four selected isolates a partial 16S rDNA sequence consisting of 869–883 nucleotides was determined by BCCMTM/LMG, University of Ghent, Belgium, as previously described (Vancanneyt et al. 2001).
Chemical analyses. Initially and at the time of sensory rejection salt content and percentage dry matter was determined (Jørgensen et al. 2000). Drip loss, pH, total volatile nitrogen (TVN), TMAO and trimethylamine (TMA) were determined as previously described (Dalgaard et al. 1993). Formation of biogene amines and organic acids were measured at regular intervals using previously described HPLC methods (Dalgaard and Jørgensen 2000; Jørgensen et al. 2000). External standards were used for identification and quantification of compounds. Gas composition within packs was measured using a Combi Check 9800-1 gas analyser (PBI, Dansensor, Ringsted, Denmark).
Sensory analyses. Five to six trained panellists carried out sensory evaluations. Samples were coded with 3-digit numbers and served in randomized order to the panellists. Prior to serving, the samples were placed at 10°C for 30 min to allow the temperature of the samples to be the same. Each panellist evaluated two portions of shrimps from each sub-batch at each time of analysis. At each sensory session, samples of freshly thawed shrimps were included to reduce the risk of panellists guessing the development in sensory scores. Changes in overall acceptance (appearance, texture, smell and taste) of the samples were evaluated by using a simple three-class scale (I, II or III) with class III corresponding to sensory rejection (Dalgaard 2000). Time of sensory rejection was defined as the time when 50% of the panellists evaluated samples from a sub-batch to be in class III. Furthermore the panellists were asked to describe the sensory characteristics of the samples using a predefined vocabulary and/or their own words.
Spoilage activity of microbial isolates. Shrimps were inoculated with mixtures of strains from each of the dominating groups of micro-organisms found on the naturally contaminated shrimps at the time of sensory rejection. Each mixture consisted of four strains precultured individually (5°C, 1–2 d) in APT broth (Difco 265510) or tryptone soya broth (Oxoid CM129), respectively, for Gram-positive and Gram-negative isolates. Precultures were harvested late in their exponential growth phase, defined as an increase in absorbance of 0·05–0·5 at 540 nm (Novaspec II, Pharmacia LKB Biochrom, Cambridge, UK). Inoculation mixtures were prepared by mixing the four precultures followed by a dilution in PS to a cell density of 107 CFU ml−1. Shrimps were added 1% (v/w) of the relevant mixture(s) to achieve an initial concentration of 105 CFU g−1. Inoculation mixtures were added as four portions (4 × 0·25%, v/v) and after each addition the shrimps were manually tumbled to ensure an even distribution of micro-organisms on the samples. The samples were packed as previously described for the storage trials and stored at 5°C. Three packs from each sub-batch were analysed at the day of packaging and after 10 d using the following analyses previously described for the storage trials: LH, NAP, STAA and sensory evaluation.
Shrimps were inoculated with a mixture of four strains to an initial level of approx. 102L. monocytogenes per gram. The four strains, originally isolated from seafood (Jørgensen and Huss 1998) were precultured, mixed and inoculated on shrimps as described above under the section Spoilage activity of microbial isolates. Two experiments (April–May 2003 and August–September 2003) each consisting of three sub-batches were carried out as previously described for the storage trials. In the first experiment the three sub-batches were stored at 2, 5 and 8°C just after packaging whereas in the second experiment all three sub-batches of the inoculated MAP shrimps were stored frozen at approx. −22°C during 120 d (April–July 2003) and then stored at 2, 5 and 8°C respectively. The temperature of all sub-batches of shrimps was recorded throughout the challenge tests as described for storage trials above.
Sampling and analyses. At regular intervals, three packs from each sub-batch were analysed. Samples were homogenized and diluted as described above for the storage trials. Listeria monocytogenes was enumerated by spread plating (37°C, 48 h) on Palcam agar (CM877, Oxoid) supplemented with Palcam Selective Supplement (SR0150). LAB was enumerated using NAP agar to examine the importance of LAB on growth of L. monocytogenes. To study the effect of frozen storage on inactivation of L. monocytogenes during storage at −22°C three samples were, at regular intervals, removed during the frozen storage period and analysed as described above.
Evaluation of models to predict microbial growth in MAP shrimps
To determine lag time (h), maximum specific growth rate (μmax, h−1) and maximum population density (MPD, log CFU g−1), the four-parameter logistic model (Dalgaard 1995) were fitted to growth data determined on NAP in storage trials and on NAP and Palcam agar in challenge tests. The software package Fig.P (v. 2·98, Biosoft, Cambridge, UK) was used for curve fitting. An F-test to compare fits of the three- and four-parameter logistic models was used to evaluate whether lag phases of the microbial growth curves were significant (Dalgaard 1995). One-way anova and multifactor anova were used to determine whether differences between mean values of repeated measurements were statistically significant. Calculations were carried out using Statgraphics Plus (Anon. 1998). A simple model for the inhibitory effect of LAB on growth of L. monocytogenes (Giménez and Dalgaard 2004) was evaluated with data from the cooked and peeled MAP shrimps. In addition, μmax-values observed for L. monocytogenes in challenge tests were compared with available predictive models (Augustin and Carlier 2000; Devlieghere et al. 2001; Growth Predictor (IFR 2003)). Observed and predicted μmax-values were compared by calculation of bias- and accuracy-factors (Ross 1996).
Time to toxin production by Cl. botulinum type E was predicted by using both the Growth Predictor (IFR 2003) and the Pathogen Modelling Program (USDA 2004).
Product characteristics and storage conditions. The shrimps studied had an initial pH of 7·7 ± 0·0 and contained 1·91 ± 0·02% water phase salt, 20·0 ± 1·2% dry matters, 10·3 ± 0·2 mg TVN 100 g−1 and 599 ± 52 ppm of water phase lactic acid. After frozen storage at −22°C the thawed MAP shrimps reached their chill storage temperatures of 2, 5 and 8°C, after c. 4·6, 2·7 and 0·4 d respectively. Chill storage temperatures were 1·8 ± 0·3, 5·0 ± 0·4 and 8·0 ± 0·2°C in the experiment without previous frozen storage and 1·7 ± 0·4, 5·0 ± 0·3 and 7·9 ± 0·3°C for the thawed MAP shrimps. Average temperature during the frozen storage period (120 d) was −21·7 ± 0·5°C. The CO2 and O2 concentration in the modified atmosphere measured immediately after packaging was 44·7 ± 1·8 and 22·5 ± 0·7%, respectively. The CO2 concentration decreased rapidly after packaging due to absorption in the shrimp flesh and equilibrium levels of 24·2 ± 1·2, 24·6 ± 1·1 and 26·0 ± 1·4% CO2 were reached at 2, 5 and 8°C respectively. After frozen storage the equilibrium CO2 concentrations were 25·1 ± 1·0, 24·8 ± 2·1 and 26·4 ± 1·5% at 2, 5 and 8°C, respectively, for the thawed MAP shrimps. The gas volume in each pack decreased when CO2 dissolved in the shrimps and this caused the concentration of O2 in the modified atmosphere to increase to c. 30%. At the end of the storage period the concentration of CO2 and O2 increased and decreased respectively.
Microbiological changes. Four months frozen storage at −21·7 ± 0·5°C significantly reduced (P = 0·013) the initial concentration of micro-organisms (APC) in MAP shrimps from 2·70 ± 0·0 log (CFU g−1) to 2·43 ± 0·11 log (CFU g−1). In the MAP shrimps without frozen storage the initial concentration of LAB was 1·30 ± 0·0 log (CFU g−1) corresponding to c. 5% of APC but in the frozen and thawed MAP shrimps LAB was not detectable, by the method used, at the first time of sampling after thawing (Fig. 1). Nevertheless, LAB always dominated the spoilage microflora of the MAP shrimps and frozen storage had no significant effect on their maximum specific growth rates (μmax) and maximum cell densities (P = 0·09–0·57) (Table 1, Fig. 1). In addition to LAB, presumptive B. thermosphacta growing on STAA agar constituted a considerable part of the spoilage microflora (Table 1) and B. thermosphacta was enumerated at regular intervals during storage of the frozen and thawed MAP shrimps (Fig. 1).
|LH (log CFU g−1)||8·3 ± 0·1||8·6 ± 0·1||8·6 ± 0·1||7·7 ± 0·1||8·2 ± 0·2||8·5 ± 0·1|
|NAP (log CFU g−1)||8·1 ± 0·2||8·6 ± 0·2||8·6 ± 0·1||7·7 ± 0·2||8·1 ± 0·1||8·4 ± 0·1|
|STAA (log CFU g−1)||7·8 ± 0·1||8·3 ± 0·0||8·3 ± 0·2||7·5 ± 0·1||7·8 ± 0·2||8·0 ± 0·1|
|IA-H2S (log CFU g−1)||6·2 ± 0·2||6·4 ± 0·6||6·7 ± 0·5||5·3 ± 0·2||5·1 ± 1·2||5·3 ± 0·2|
|Rogosa (log CFU g−1)||<2||<2||<2||<2||<2||<2|
|Slanetz & Bartley (log CFU g−1)||<2||<2||<2||<2||<2||<2|
|VRBG (log CFU g−1)||<2||NT||NT||<2||2·9 ± 0·2||3·8 ± 0·2|
|Drip loss (%)*||18·1 ± 1·9||17·1 ± 2·2||19·7 ± 1·6||18·7 ± 0·9||19·9 ± 1·9||18·1 ± 1·1|
|pH||7·5 ± 0·0||7·8 ± 0·1||7·6 ± 0·0||7·6 ± 0·0||7·6 ± 0·1||7·5 ± 0·1|
|TVN (mg-N/100 g)||19·6 ± 3·6||17·9 ± 9·5||16·6 ± 6·7||11·7 ± 2·0||14·2 ± 5·0||16·2 ± 2·9|
|Water phase lactic acid (ppm)||488 ± 15||511 ± 18||516 ± 17||640 ± 12||596 ± 71||560 ± 84|
|Shelf life (d)||25–26||15–16||9–10||25–26||14–15||9–10|
|Sensory spoilage characteristics||Sour, wet-dog and chlorine-like||Sour, wet-dog and chlorine-like|
Of 116 isolates from the dominating spoilage microflora of MAP shrimps 60% were identified as C. maltaromaticum, 27% as B. thermosphacta and 13% as Psychrobacter spp. (Table 2). Listeria monocytogenes was not detected in shrimps that had not been inoculated (results not shown).
|Total number of isolates||19 (100)||20 (100)||19 (100)||19 (100)||19 (100)||20 (100)|
|Gram-positive||17 (89)||18 (90)||15 (79)||16 (84)||17 (89)||18 (90)|
|Carnobacterium maltaromaticum*||11 (58)||12 (60)||11 (58)||11 (58)||12 (63)||13 (65)|
|Brochothrix thermosphacta†||6 (31)||6 (30)||4 (21)||5 (26)||5 (26)||5 (25)|
|Gram-negative||2 (11)||2 (10)||4 (21)||3 (16)||2 (11)||2 (10)|
|Psychrobacter spp.‡||2 (11)||2 (10)||4 (21)||3 (16)||2 (11)||2 (10)|
Chemical and sensory changes. At the time of sensory spoilage of the MAP shrimps, drip losses of 17·1 to 19·9% were observed but no significant effect of previous frozen storage or chill storage temperature was observed (P = 0·32–0·76) (Table 1). Compared with the initial concentration of TVN (10·3 ± 0·2 mg-N 100 g−1) relatively small amounts were formed during storage of the MAP shrimps (Table 1). At 2°C, previous frozen storage, significantly lowered (P < 0·0001) the formation of TVN but this was not observed at 5 and 8°C (P = 0·25–0·87). TMA was not produced in any of the sub-batches at the time of sensory spoilage (results not shown). The concentration of lactic acid decreased during chilled storage of MAP shrimps (P = 0·017–0·047) but this was not observed for the frozen and thawed MAP shrimps (P = 0·25–0·94) an observation that needs to be further examined (Table 1). Irrespective of the treatment, no organic acids or biogenic amines were formed in the MAP shrimps at the time of spoilage. Previous frozen storage had no effect on shelf life and sensory spoilage characteristics of the cooked and peeled MAP shrimps (Table 1)
Spoilage activity of microbial isolates. A mixture of C. maltaromaticum and B. thermosphacta isolates produced the same off-flavours in cooked and peeled MAP shrimps as those observed in the naturally contaminated spoiled product (Tables 1 and 3). Brochothrix thermosphacta alone produced strong buttermilk-like and sour off-odours but these clearly differed from the sensory spoilage characteristics of the naturally contaminated product. Alone C. maltaromaticum and Psychrobacter spp. produced only weak off-flavours (Table 3).
|Control||3·5 ± 0·1||I||None||6·9 ± 0·3||I||None|
|Carnobacterium maltaromaticum||5·4 ± 0·3||I||None||9·0 ± 0·2||II||Chlorine-like|
|Brochothrix thermosphacta||5·5 ± 0·1||I||None||8·6 ± 0·1||III||Buttermilk-like, sour|
|C. maltaromaticum in co-culture with B. thermosphacta||5·5 ± 0·3||I||None||8·8 ± 0·1||III||Wet dog, chlorine-like, sour|
|Psychrobacter spp.||5·2 ± 0·5||I||None||8·8 ± 0·1||II||Fishy (slightly)|
Storage and product characteristics. Shrimps used for the challenge test had product characteristics as described previously for the storage trial. The three sub-batches without preceding frozen storage were stored at 1·7 ± 0·1, 4·6 ± 0·5 and 7·8 ± 0·2°C respectively. The average temperature during the frozen storage period was −21·2 ± 1·1°C and after thawing MAP shrimps were kept at 1·6 ± 0·9, 4·6 ± 0·5 and 8·2 ± 0·2°C.
Microbiological changes. The concentration of L. monocytogenes after inoculation of the MAP shrimps was 1·5 ± 0·3 log (CFU g−1) and this level was not influenced by 120 d of frozen storage at −22°C (Fig. 2). At all the conditions tested L. monocytogenes grew without significant lag phases (Table 4) and the frozen storage period had no significant effect on its maximum specific growth rate (P = 0·10–0·29). However, the time for a 100-fold increase in the cell concentration of L. monocytogenes was increased by previous frozen storage (Table 4). The MPD of L. monocytogenes increased with the storage temperature and previous frozen storage significantly lowered the MPD values at 2°C (P = 0·0064) and 5°C (P = 0·0026) but not at 8°C (P = 0·27) (Table 4). In the challenge test, LAB that occurred naturally in the MAP shrimps grew 3·3 to 4·3 times faster than L. monocytogenes at 2°C and LAB always reached MPD of 8·6–9·0 log (CFU g−1) (Table 4).
|Lag phase (h)||NS||NS||NS||NS||NS||NS|
|MPD (log CFU g−1)||5·2||6·0||7·2||4·5||5·3||7·2|
|MPD, predicted by interaction model (log CFU g−1)||3·6||6·2||5·6||3·4||5·1||6·0|
|Time to 1000-fold increase (d)||33·3||12·2||6·3||42·1||16·0||7·6|
|Time to 100-fold increase (d)||20·3||8·5||4·3||30·6||11·7||5·7|
|Time to 10-fold increase (d)||7·4||4·6||2·3||18·8||7·4||3·8|
|Lactic acid bacteria|
|Lag phase (h)||138||45||NS||ND||ND||ND|
|MPD (log CFU g−1)||8·6||8·8||8·6||9·0||8·8||8·7|
|μmax ratio (LAB/L. monocytogenes)||4·3||1·8||1·7||3·3||1·8||1·6|
Evaluation of models to predict microbial growth in MAP shrimps
The interaction model for the inhibitory effect of LAB on L. monocytogenes predicted MPD appropriately at 5°C but at 2 and 8°C MPD of L. monocytogenes were underestimated (Table 4). Maximum specific growth rates (μmax) predicted with available models for L. monocytogenes were on average 28–32% higher than the μmax-values observed in the present study with cooked MAP shrimps at 2, 5 and 8°C (Table 5). Time to toxin production by nonproteolytic Cl. botulinum at 5°C and 8°C was predicted to be 25 and 12 d, respectively, using both the Growth Predictor software (IFR 2003) and the Pathogen Modelling Program (USDA 2004). At 2°C toxin was predicted not to be formed.
|Augustin and Carlier (2000)||−2·7–45·5||0–164†||0–12·6||0–3·2||4·5–9·6||1·32||1·32|
|Devlieghere et al. (2001)||4–12||0–80||2·0–6·2||0–3||6·2||1·28||1·28|
|Growth Predictor v. 1·01‡||1–35||0–100||0–11·4||–||4·4–7·5||1·31||1·31|
Potential growth of L. monocytogenes limited shelf life of cooked and peeled MAP shrimps more than growth of spoilage bacteria and nonproteolytic Cl. botulinum (Tables 1 and 4). To prevent growth of L. monocytogenes to critical concentrations our data suggest that cooked and peeled MAP shrimps should be kept at 2°C or colder. In addition, maximum declared shelf life, corresponding to the time required for L. monocytogenes to multiply 100-fold with no lag phase, should be limited to 20–21 d at 2°C, 8–9 d at 5°C and 4–5 d at 8°C (Table 4, Fig. 2). Listeria monocytogenes is inactivated by cooking of shrimps but the pathogen can be present in the cooked and peeled product (Destro et al. 1996; Valdimarsson et al. 1998) and limits to storage temperature and shelf life, as just indicated, are therefore important for this RTE product.
Frozen storage (120 d at −22°C) extended the time for a 100-fold increase in cell concentration of L. monocytogenes by 10·3, 3·2 and 1·4 d at 2, 5 and 8°C respectively (Table 4). This delay in growth was longer than the c. 4·6, 2·7 and 0·4 d required for shrimps to reach 2, 5 and 8°C, respectively, after frozen storage and this indicated an inhibitory effect of frozen storage on the subsequent growth of L. monocytogenes in thawed and chilled MAP shrimps, especially when stored at 2°C. Listeria monocytogenes can be inactivated by frozen storage of food with pH below 5 but the limited effect observed in the present study with shrimps (pH = 7·7) corresponds to results for other foods with pH above 5·8 and kept at −18°C (Palumbo and Williams 1991). Consequently, frozen storage cannot be used to inactivate L. monocytogenes or substantially delay growth in thawed MAP shrimps.
The storage temperature had a pronounced effect on both shelf life, determined by sensory evaluation, and growth of L. monocytogenes in the cooked and peeled MAP shrimps (Tables 1 and 4). In fact, the effect of temperature on shelf life was more pronounced than predicted by the classical square-root spoilage model for fresh fish from temperate waters as also previously observed for cooked and brined MAP shrimps (Dalgaard and Jørgensen 2000). The shelf life of the cooked and peeled shrimps was appropriately described by the exponential spoilage model when using a slope parameter of 0·16°C−1 (http://www.dfu.min.dk/micro/SSSP). This high temperature sensitivity most likely results from inactivation by cooking of the natural psychrotolerant microflora on shrimps. For L. monocytogenes growth at 2°C was markedly slower than at both 5 and 8°C (Table 4). This may be related to a high antilisterial effect of CO2 at 2°C. The equilibrium concentrations in the modified atmosphere indicated only a slightly increased solubility of CO2 at 2°C as compared with 5 and 8°C and this was expected based on the temperature, initial CO2 concentration and the gas/product ratio (Ross and Dalgaard 2004, pp. 129–131). Claire et al. (2004) observed a marked inhibiting effect of MAP (40% CO2/60 N2 and 80% CO2/20% N2) on growth of L. monocytogenes in hard-boiled eggs (pH = 7·7) at 4°C but not at 8 and 12°C. However, the present study cannot determine whether a combined effect of temperature and CO2 or simply the low temperature of 2°C which is close to the growth limit of L. monocytogenes accounted for its slow growth in cooked MAP shrimps at 2°C.
The growth of L. monocytogenes were inhibited in cooked and peeled MAP shrimps by high concentrations of LAB and this effect was most pronounced at 2°C where the difference in growth rates between LAB and L. monocytogenes was largest (Fig. 1, Table 4). The simple interaction model to predict the growth-inhibiting effect of LAB on L. monocytogenes on average overestimated the inhibiting effect of LAB (Table 4). The model was developed and validated for cold-smoked salmon (Giménez and Dalgaard 2004) and this could explain our data as the inhibitory effect of bacteriocins, produced by LAB, against L. monocytogenes can be stimulated by NaCl in concentrations found in cold-smoked salmon (Lebois et al. 2004). In cooked MAP shrimps the interaction of LAB and L. monocytogenes, however, seems of little practical importance as the product becomes sensory spoiled by the same concentrations of LAB required to inhibit the pathogen (Table 1, Fig. 1).
Growth of L. monocytogenes in cooked and peeled MAP shrimps was slightly overestimated by the three evaluated predictive models (Table 5). All models included the effect of temperature, CO2, NaCl and pH on growth of L. monocytogenes in ranges reflecting the conditions of the present study with cooked and peeled MAP shrimps. However, none of the evaluated models included the effect of O2 previously described as inhibiting the growth of
Cold-water prawns can be taken all year round, but fishing is limited by the ice conditions, especially in Ilulissat when the Disko Bay freezes. The prawns are caught both inshore and offshore.
Inshore prawn fisheries
Inshore, the cold-water prawns are caught by local fishermen in small vessels. They fish quite close to the shore, but at great depths, to reach the prawns’ seabed habitat. The prawns are caught in a trawl net and hauled on board. Then the prawns are stored on ice for a maximum of four days, or until the fishermen have filled their on-board storage facilities. The catch is landed at one of Royal Greenland's factories, where the prawns’ size and appearance are rated prior to processing.
In Greenland, all catches of inshore prawns are cooked and peeled in the land-based factories.
Offshore prawn fisheries
Offshore fishing for cold-water prawns is also by trawl, but using much larger vessels and trawls. Royal Greenland operates two seagoing trawlers, Akamalik and Qaqqatsiaq, which fish solely for prawns. The trawl is lowered several hundred metres onto the seabed, where the prawns are trapped in the net and hauled on board for processing.
Royal Greenland's trawls are fitted with sorting grids and escape panels that lead any larger fish that have entered the trawl opening back into the sea. Royal Greenland also supports the research and development of less invasive trawling methods with less impact on the seabed and reduced fuel consumption for the fishing vessels.
Read more about Royal Greenland's sustainable fisheries strategies here
In Greenland, it is mandatory to land minimum 25 % of the total catch for on-land processing. Our trawlers unload their catch of prawns to one of our land-based prawn processing factories in Ilulissat or Sisimiut, where minimum 25 % of the catch is processed into cooked and peeled prawns.
The fisheries in Western Greenland, Canada and Norway are all certified as sustainable by the Marine Stewardship Council and evaluated to be well-managed and supervised via mandatory catch logs.
Read more about MSC certification here.
See how the finest prawns are caught in the Disko Bay area in western Greenland.Time:
See how the finest shell-on prawns are caught by our seagoing factory trawlers, far from shore in the North Atlantic and Arctic oceans.Time: