Shelf life extension of whole Norway lobster Nephrops norvegicus using modified atmosphere packaging

International Journal of Food Microbiology 167 (2013) 369–377

Contents lists available at ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Shelf life extension of whole Norway lobster Nephrops norvegicus using modified atmosphere packaging Sebastian G. Gornik a,1, Amaya Albalat a,2, Chonchanok Theethakaew b,3, Douglas M. Neil a,⁎ a b

Institute of Biodiversity, Animal Health and Comparative Medicine, College of Medical, Veterinary & Life Science, University of Glasgow, Glasgow, UK Institute of Infection, Immunity and Inflammation, College of Medical, Veterinary and Life Science, University of Glasgow, Glasgow, UK

a r t i c l e

i n f o

Article history: Received 8 June 2013 Received in revised form 2 October 2013 Accepted 3 October 2013 Available online 12 October 2013 Keywords: Norway lobster Modified atmosphere packaging QIM SSO Photobacterium phosphoreum

a b s t r a c t Once a nuisance by-catch, today the Norway lobster (Nephrops norvegicus) is a valuable UK fisheries commodity. Unfortunately, the species is very susceptible to quality deterioration post harvest as it quickly develops black spots and also spoils rapidly due to bacterial growth. Treatment with chemicals can stop the blackening and carefully monitored cold storage can result in a sensory shelf life of up to 6.5 days. The high susceptibility to spoilage greatly restricts the extent to which N. norvegicus can be distributed to retailers and displayed for sale. The application of modified atmosphere (MA) could be extremely beneficial, allowing the chilled product to stay fresh for a long period of time, thus ensuring higher sales. In the present study, we identified a gas mix for the MA packaging (MAP) of whole N. norvegicus lobster into 200 g retail packs. Our results show that a shelf life extension to 13 days can be achieved when retail packs are stored in MAP at 1 °C. Effectiveness of the MAP was evaluated by using a newly developed QIM for MA-packaged whole N. norvegicus and also by analyzing bacterial plate counts. Changes in the microflora and effects of different storage temperatures on the quality of the MA packs are also presented. The main specific spoilage organism (SSO) of modified atmosphere packaged Norway lobster is Photobacterium phosphoreum. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Once considered a nuisance by-catch, currently the Norway lobster (Nephrops norvegicus) represents one of the most valuable fisheries in the UK (Neil, 2012) and whole chilled animals are highly priced. Unfortunately however, the species is very susceptible to spoilage and exhibits rapid quality deterioration post-harvest (Gornik et al., 2011; Neil, 2012), especially when icing and/or cooling are delayed (Albalat et al., 2011). Furthermore, when left at elevated temperatures N. norvegicus also develops brown–black spots and lesions Neil, 2012 (Martínez-Álvarez et al., 2007) quickly, due to the accumulation of melanin triggered by the enzyme polyphenol oxidase (PPO) (Giménez et al., 2010). This browning (melanosis) is further accelerated after trawl capture and rough handling (Bartolo and Birk, 2002). Therefore, to inhibit melanosis the N. norvegicus processing industry routinely uses different anti-melanotic formulations (i.e. metabisulfite, 4hexylresorcinol based products). One processing method that is commonly used to extend the shelf life of seafood is modified atmosphere packaging (MAP) and, based on ⁎ Corresponding author. Tel.: +44 141 330 5975. E-mail address: [email protected] (D.M. Neil). 1 Present address: School of Botany, University of Melbourne, 3010 Victoria, Australia. 2 Present address: Institute of Aquaculture, School of Natural Sciences, University of Stirling, Stirling, UK. 3 Present address: Department of Microbiology, Faculty of Public Health, Mahidol University, Bangkok, Thailand. 0168-1605/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijfoodmicro.2013.10.002

data from fish and other shellfish, a general shelf life increase of 30–60% is common (Sivertsvik and Jeksrud, 2002). Unfortunately, studies investigating the effects of MAP in crustaceans are sparse with shrimps and prawns being the most studied products. To date only two gas mix conditions have been tested for MAP application in N. norvegicus (Ruiz-Capillas et al., 2003). Unfortunately, that study only assessed these two gas mixes for bulk storage of lobster at 1 °C prior to sale, but not for MAP in retail packs. Freshness and changes in sensory attributes are critical parameters for the quality assessment of seafood and form the basis of many quality index methods (QIMs) (Bremner, 1985; Hyldig et al., 2007), and as such they are also important for the assessment of N. norvegicus (Albalat et al., 2011; Neil, 2012). Numerous QIM schemes for various types of fish and crustaceans exist (e.g. Huidobro et al., 2000; Barbosa and Vaz-Pires, 2004; Cardenas Bonilla et al., 2007). Gómez-Guillén et al. (2007) and Martínez-Álvarez et al. (2008) have published QIMs for whole N. norvegicus treated with an anti-melanotic. However, these QIMs were only tested and optimized for air storage at 2 °C and 4 °C, respectively. Dependency on storage temperature and suitability for MA packs have not been tested for these QIMs. The aim of the present study was to provide an optimal MAP gas mix and QIM for MA-packaged whole N. norvegicus lobster after treatment with an anti-melanotic stored within retail packs. Firstly, a QIM for whole N. norvegicus was tested and calibrated. The QIM scheme was then modified to work with animals that had been treated with an anti-melanotic. Then, a suitable gas mix for MAP of N. norvegicus was

370

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

identified using a triangular, non-biased approach. At the same time the whole animal QIM was adapted for MA packs. The quality of the product stored in the MA packs was further evaluated using previously established microbiological methods. Changes in the microflora and effects of different storage temperatures on the quality of the MA packs were also assessed. 2. Material and methods 2.1. Trawl capture and post catch storage of whole N. norvegicus N. norvegicus were caught in the Clyde Sea area, Scotland, UK (55°36.65′N and 04°52.99′W) as described by Gornik et al. (2011). Once caught, the lobsters were thoroughly washed for 5 min in a fishing basket using seawater derived from an on-board high-pressure hose, following common practice in the industry. Within 30min (upon arrival at the pier) approximately 20 kg of lobsters (or around 500 animals) was placed in a fish box and 10 kg of ice was placed on top of the content. The box was then stored in a cold room at 5 °C until further processing, which was approximately 24 h later to simulate commercial practices. 2.2. QIM development and assessment procedure The QIM for whole N. norvegicus lobsters consists of five different parameters and four demerit scores (from 0 to 3) per parameter. The detailed descriptions of the parameters and the demerit quality scores can be found in Supplementary Table S1. The QIM was firstly calibrated for iced storage. During calibration, per data point, three assessors sampled the product (10 animals per condition) independently. In all cases, prior to assessment, the samples were equilibrated to room temperature for 15min. To avoid bias, sample identity was not disclosed until after evaluation. 2.3. Anti-melanotic treatment All treatments were carried out at 5 °C. After overnight storage (approx. 12 h) whole animals were dipped into a 2% solution of Melacide® SC20 (Norkem, Cheshire, UK) in seawater for 15 min following manufacturer's instructions. The animal-to-solution ratio was 1:2 (w/v). After dipping, the animals were placed in a standard fish basket to drain off any remaining liquid. 2.4. MA-packaging of whole N. norvegicus Following anti-melanotic treatment approximately 200 g of N. norvegicus (5–6 animals) was placed into Clearfresh® MA packs (R15–45, 260 × 177 × 45 mm, LINPAC, Yorkshire, UK) and packed using a PA1200 MA-packaging machine (Packaging Automation Ltd, Cheshire, UK). Gas mixes were delivered to the packaging machine through a WITT 3800 gas mixer (WITT, Cheshire, UK) fed by gas cylinders containing food-grade N2, CO2 and O2 (BOC, Scotland, UK). The CO2 was preheated to ambient temperature (5 °C) using a heated valve unit (BOC). The sealing film was anti-fog Esterpeel PS2+ AF with a transmission rate of 8 cm3 O2 per m2 day−1. After packaging the headspaces of several packs were analyzed to verify the gas mixes using a WITT OXYBABY hand-held gas analyzer. Prior to sampling the gas content of the packs was measured to ensure that packs did not have leaks during storage. Leaky packs were always discarded and not analyzed.

caught, stored on ice for 24 h and packed on-site following antimelanotic treatment. After packaging the samples were stored at 2 °C for 5 days. Then, after 5 days, the packs were transferred at 6 °C (mimicking consumer storage behaviors) for QIM sampling on day 8. See also Supplementary Fig. S1 for a schematic overview of the sampling regime. During the evaluation of the OGM-MA-packs (see below for description), a 1-step and a 3-step temperature storage regime were employed to refine the analysis (Fig. S1). In the 1-step regime (intended to serve as a ‘best practice’ positive control) the MA packs were held at a constant temperature of 1 °C. In the 3-step regime (mimicking the current commercial retail chain from processor to consumer) the samples were stored at 1 °C for 3 days (“processor”), transferred at 4 °C for 2 days (“market”) and finally stored at 6 °C for the remainder of the time (“consumer”). Temperature profiles were recorded using StowAway TidbiT® V2 (Onset, Massachusetts, USA) temperature loggers accompanying the stored packs. An excerpt of a typical temperature recording can be seen in Supplementary Fig. S2.

2.6. Microbiological analysis 2.6.1. Sample extraction and homogenization Tail meat was sampled aseptically as described by Gornik et al. (2011). Briefly, the packs (3 packs per condition/sampling time) were opened and immediately 0.5–1.0 g of tail muscle was carefully removed per tail. Up to 5 samples were taken per pack. The extracted samples were transferred into a ‘stomacher’ bag. Per 1 g of meat, 9 ml of sterile seawater containing 0.1% bacteriological peptone (Difco/BD, Oxford, UK) was added. The samples were then homogenized and transferred into sterile plastic universals for further usage.

2.6.2. Compact Dry® plates To perform high-throughput total bacterial counts (TBCs) during the developmental stage of the MAP process, Compact Dry® plates (HyServe, Uffing, Germany) were used following manufacturer's instructions. If not otherwise stated, 3 packs were analyzed per time point using 5 Compact Dry® plates per pack (3 × 5 samples).

2.6.3. MIA, IA and CFC plates To obtain more accurate total TBCs as well as the number of H2S-producing and luminous bacteria marine iron agar (MIA) plates were used and prepared following the method of Gram et al. (1987) with minor modifications (Gram et al., 1987). Instead of iron agar (IA) a marine broth (Difco/BD, Oxford, UK) was used and supplemented with 0.04% (w/v) L-cysteine (Sigma C8755; Sigma-Aldrich, Gillingham, Dorset, UK) and 0.03% (w/v) sodium thiosulfate (Sigma S7026). Pseudomonas species agar base (Oxoid, Basingstoke, Hampshire, UK) mixed with Cetrimide–Fucidin–Cephalosporin (CFC) supplement (Oxoid) was used according to the manufacturer's instructions to determine Pseudomonas counts in the samples. A series of 10-fold dilutions of the muscle homogenate (up to 10 − 6) were prepared and 100 μl of each diluent was spread onto the various culture plates and incubated at 20 °C for 48 h, when colony-forming units were counted. Luminous bacterial numbers were determined by counting plates in the dark. If not otherwise stated, per sample point 3 MA packs were analyzed using 5 plates per pack (3 × 5 samples).

2.7. Statistical analysis 2.5. Sample storage and sampling regime During preliminary trials and during the identification of a suitable gas mix, a 24-h delay was incorporated into the packaging procedure, mimicking commercial fishing practice. All the animals were freshly

The software package SPSS v.15.0 was used for statistical analysis. A Tukey's range test was used to determine statistical differences between sample means and p-valuesb0.05 were considered statistically significantly different.

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

371

2.8. Molecular analysis and phylogenetics of the microflora

3.2. Effect of anti-melanotic treatment on the QIM

Microbiota were analyzed using 16S rRNA PCR amplification using the forward primer 27F (AGAGTTTGATCMTGGCTCAG) and the reverse primer 1100R (GGGTTGCGCTCGTTG) and as described by Gornik et al. (2011). Briefly, per treatment group a total of 25 bacterial isolates were randomly picked from 5 non-selective MIA plates. Isolates were subjected to 16S rRNA PCR following DNA extraction using the boiling method. Following PCR, the amplicons were cloned into pGEM-T easy. A 658 bp region of the amplicons was bidirectionally sequenced using the primers 685R (TCTACGC ATTTCACYGCTAC) and 27F (see above). Sequence chromatograms were analyzed using the bioinformatics workflow Geneious Pro Version 5.6.3 (Biomatters, Auckland, New Zealand). ClustalW within geneious was used to align 571 bp of the nucleotide sequences of the 16S rRNA genes of the unknown bacterial isolates with those of the corresponding genes from bacterial type strains deposited at GenBank (http://www.ncbi.nlm.nih.gov/Genbank/). The sequences of the unknown bacterial isolates can be found in Supplementary File S1. The following bacterial type strains were utilized to infer species affiliation: Vibrio fischeri (ATCC 7744; accession number AY341436), Vibrio logei (ATCC 29985; EU221273), Vibrio splendidus (ATCC 33125; AF413024), Vibrio wodanis (NVI 88/441T; NR_028881), Photobacterium phosphoreum (ATCC 11040; D25310), Photobacterium profundum (DSJ4; D21226), Shewanella putrefaciens (R1418; AB208055), Pseudoalteromonas haloplanktis (ATCC 14393;·67024) and Psychrobacter nivimaris (88/2–7; NR_028948). P. nivimaris was used as an out-group. Neighborhood-joining trees were calculated within Geneious Pro Version 5.6.3 (Biomatters) using PHYML employing the Jukes–Cantor (JC69) substitution model. Only unique sequences were used for phylogenetic analysis. The clade affiliation of any unknown bacterial isolate was inferred using its relative position in relation to internal tree branches and clustering with any given type strains.

Whole N. norvegicus lobsters were treated with the anti-melanotic agent Melacide® SC20 to prevent melanosis before storage. The QIM was then scored based on the previously calibrated QIM. Per data point two assessors sampled a total of 12 packs containing 200 g of whole N. norvegicus (5–7 animals) independently. The results clearly show that the QIM scores changed due to Melacide® SC20 treatment (Fig. 1A). Although the samples were still rejected from 6.5day onwards, the QIM score did not reach the same values (b 8) when compared to untreated whole lobsters at the same time point (N8). Only after 9 days was a score equal to the rejection score of untreated samples recorded. When the QIM parameters were analyzed individually it was determined that only two parameters, namely ‘head’ and ‘odor’, influenced the overall score significantly (Supplementary Fig. S3). All other parameters (‘claws’, ‘upper tail’ and ‘underside tail’) lost resolution between day 5 and day 7 and did not contribute further to the QIM score after this time point. Barbosa and Vaz-Pires (2004) showed that if QIM parameters stop changing significantly over time, they become meaningless and must be omitted from the QIM scheme. Therefore, for the QIM of N. norvegicus treated with an anti-melanotic the parameters ‘claws’, ‘upper tail’ and ‘underside tail’ were removed. Furthermore, since the changes in the odor were more significant than the changes in the quality and appearance of the ‘head’ region, the ‘head’ parameter was also omitted. Based on the 6.5-day rejection point of the untreated whole animal QIM, the point of rejection for whole N. norvegicus lobsters treated with the anti-melanotic was determined to occur at a score of above 2.0 (Fig. 1B). At this score the odor of the sample was described as ‘old seaweed’, ‘musty’ and ‘slight ammonia’. The updated odorbased QIM scheme for whole N. norvegicus lobsters treated with an anti-melanotic can be found in Supplementary Table S2.

3.3. Identification of a suitable gas-mix for whole N. norvegicus lobsters 3. Results

3.0

9

8

ay D

D

ay

7

6

ay D

5

ay D

ay D

3

ay D

0.0

2

0

ay

0.5

Fr es h D ay 1 D ay 2 D ay 3 D ay 4 D ay 5 D ay 6 D ay 7 D ay 8 D ay 9

2

4

1.0

D

4

1.5

1

6

2.0

ay

8

n = 12

D

10

QIM Score

2.5

B

h

n = 12

QIM Score [odour]

12

ay

A

es

14

D

A scatter plot of the QIM scores shows a high correlation between the QIM score and the storage time on ice (Fig. 1A). Furthermore, it was evident that the overall quality of the samples deteriorated over time. If the sample reached scores over 8 the product was rejected by the panelists and rendered unfit for consumption due to severe melanosis and putrid fishy odors. This rejection occurred after an average of 6.5 days.

Due to the very limited information for MAP of N. norvegicus, a triangular un-biased approach was chosen to find the best suitable gas-mix for the MA-packaging of this species. Twenty gas mixes were tested initially (Fig. 2). For QIM, per data point a minimum of two assessors independently sampled a total of 3 packs. On day 5, when the MA packs were first opened, the QIM score of most packs was significantly lower than the air control (Fig. 3A). The air control, as well as several mixes, exceeded the 2.0 score QIM limit and was discarded. By day 8, more gas mixes reached the QIM limit of 2.0, but some scored below this limit and were therefore regarded as potential

Fr

3.1. Calibration of a whole N. norvegicus lobster QIM

Fig. 1. QIM results. (A) Black circles show the QIM scores for whole N. norvegicus during storage on ice. The star depicts the point of rejection due to off-odors as determined previously by Gornik et al. (2011). White triangles show the QIM scores for whole lobsters treated with the anti-melanotic compound Melacide SC20. (B) Black triangles show the QIM scores for the odor component of the QIM only. The star depicts the point of rejection. Each point represents the mean +/− the standard error of the mean (SEM) of 12 samples.

372

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

candidates for MA-packaging of whole N. norvegicus lobsters (Fig. 3B). These mixes all had a high CO2 proportion (above 45%). The samples were also analyzed for their TBC using Compact Dry TC plates. Using 5 Compact Dry TC plates per pack, a total of 3 packs were analyzed per data point. The results show that on day 5 the TBC for virtually all MA packs was significantly reduced compared to that of the air control (Fig. 3C). On day 8, however, only one gas mix (10:80:10; O2:CO2:N2) showed a persistent and significant bacteria reduction (Fig. 3D) while still exhibiting favorable QIM results. This gas mix was therefore chosen for MA-packaging of whole N. norvegicus lobsters and a more thorough analysis was carried out. From here onwards the 10:80:10 gas mix will be termed the ‘optimal gas mix’ (OGM). 3.4. Detailed evaluation of OGM-packaged whole N. norvegicus lobster The effects of the previously established OGM on the total bacterial count were further assessed using MIA plates, which allowed the simultaneous determination of the TBC, H2S-producing bacteria count and luminous bacteria count. In addition, CFC selective agar plates were used to investigate the number of Pseudomonas species. For the packs that were held at a constant low temperature, the changes in the microflora were analyzed and compared with changes in the air control packs using molecular techniques. Per data point 3 packs were analyzed using 5 plates per pack. 3.4.1. Bacteria counts and QIM during storage at 1 °C During storage at 1 °C the TBC was significantly lower in the MA packs compared with the air control (Fig. 4A). An initial decrease from 4.02 log10 cfu g−1 to 2.60 log10 cfu g−1 was observed after 3 days, then the TBC increased to 4.24 log10 cfu g−1 by day 7 and reached 5.09 log10 cfug−1 by day 13. In the air control, the TBC rose continuously from day 1 onwards and reached values of 7.84 log10 cfu g−1 on day 13. Similar trends were found in the counts of both the H2S-producing bacteria and the luminous bacteria. However, the counts were lower as these types of bacteria represent only a fraction of the total count, being on average an order of magnitude lower (Fig. 4B and C). In the MA packs no Pseudomonas sp. were detectable, while in the air control the Pseudomonas counts increased from 2.0 log10 cfu g−1 to 4.0 log10 cfu g−1 from day 3 onwards (Fig. 4D). 0 100

N. norvegicus that were packed in the OGM and stored at 1 °C received QIM scores of 0 until day 3, indicating a marine, fresh, hay-like odor (Fig. 4E). Between day 3 and day 5 these odor attributes disappeared and on day 5 the packs received a score of 1, having a less fresh or neutral odor. During the next 8 days the odor changed only slightly and by day 13 scores around 1 were obtained, but never reaching a score of 2. The MA packs were never rejected, nor described as releasing off-odors. The air controls, on the other hand had lost their acceptable odor attributes by day 3 (score 1) and were rejected by day 5 (score 2). By day 9 all air control packs released strong, sour, ammonialike, musty off-odors (the maximum score of 3). 3.4.2. Bacteria counts and QIM during step-temperature storage The effect of a stepped temperature regime on the OGM-MA packs was measured. The initial total bacterial counts were found to be at 3.61 log10 cfu g−1 in the fresh samples (Fig. 5A). This count decreased to 2.77 log10 cfu g−1 during the 24 h of iced storage prior to MA packaging. In the first 3 days of storage the OGM-MA packs and air controls behaved similarly and reached levels of 3.60 log10 cfu g−1, but after the temperature was raised on day 3 there was a large increase in the total bacterial count of the air control, with bacterial numbers reaching 6.55 log10 cfu g−1 on day 5. After the temperature was raised again on day 5, the TBCs in the OGM-MA packs became more similar to those in the air control (5.72 log10 cfu g−1 and 6.48 log10 cfu g−1, respectively) and remained high (7.41 log10 cfu g−1) until day 9 when the experiment was terminated due to QIM rejection. The counts of the H2S-producing bacteria and luminous bacteria showed similar trends to the total counts, with the same temperature-dependent pattern (Fig. 5B and C). Pseudomonas sp. counts stayed below the detection limit in the OGM-MA packs, while in the air control they increased continuously to reach a level of 3.82 log10 cfu g− 1 on day 9 (Fig. 5D). Whole N. norvegicus packaged in the OGM-MA and stored following the 3-step-temperature regime received scores of 0 and were described as releasing fresh, hay-like, marine odor after 3 days of storage (Fig. 5E). On day 5, after the temperature was raised to 4 °C, these odor attributes were lost and the odor was described as neutral. The packs showed no further significant change up to day 7, but on day 8 the MA packs were described as liberating old seaweed-like, musty and slightly ammonia-like odors and were rejected (score 2). The experiment was continued to day 9 when a score of 2.5 was obtained. By comparison, the air control packs received scores of above 2 by day 5 and were rejected at that time point.

10 90 20 80

[%

]

70

RB

40

ON

OG

EN

60

O DI

50 50

XID

TR

CA

30

NI

E[

60

%]

40 70 30

80 20 90 10 100 0 0

10

20

30

40

50

60

70

80

90

100

3.4.3. Changes in microflora under MA during storage at 1 °C The bacterial microflora in the tail muscle of whole N. norvegicus treated with an anti-melanotic and stored in the OGM-MA at 1 °C for 8 days was analyzed and compared with fresh samples and air controls. A higher diversity of bacterial isolates including Vibrio (82%), Photobacterium (6%), Pseudoalteromonas (6%), Shewanella (6%) and Psychrobacter (6%) species was found in fresh samples, when compared to the samples stored for 8 days (Fig. 6). In the air control packs, after 8 days of storage only, Vibrio and Photobacterium species were isolated with Vibrio species representing 45% and Photobacterium 55% of the isolated bacteria. On the other hand, in the OGM-MA packs the Photobacterium clade dominated (94%) after 8 days of storage with virtually all isolates belonging to P. phosphoreum (as indicated by the close association of 19 unknown bacterial isolates with a known P. phosphoreum species in the MAP tree in Fig. 6). Vibrio species were much less prevalent in this sample as only 6% of the isolates were found to belong to this clade.

OXYGEN [%] Fig. 2. Ternary plot of gas mixes used to identify a modified atmosphere suitable for the shelf life extension of whole, treated N. norvegicus. Black circles indicate gas mixtures of the 3 gases CO2, N2 and O2 that did not pass the QIM on day 8, while white circles indicate gas mixtures that passed the QIM.

4. Discussion The ultimate goal of this study was to determine an optimal gas mix and develop a QIM for MA-packaged whole N. norvegicus lobster

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

A

B

C

D

373

Fig. 3. QIM scores and Compact Dry TC plate bacteria counts in MA packs. (A) QIM scores of MA packs after 5 days of storage at 2 °C. All MA packs score significantly lower than the air control and remain below the rejection cut-off score. (B) QIM scores in MA packs after 8 days of storage with 5 days at 2 °C and 3 days at 6 °C. A total of 6 MAs receive scores significantly lower than the remaining MAs and stay below the rejection cut-off score of 2. (C) Total bacteria count in various MA packs after 5 days of storage at 2 °C. The bacteria counts in all MA packs are significantly lower than the bacteria count in an air control. (D) Total counts in MA packs after 8 days of storage with 5 days at 2 °C and 3 days at 6 °C. Most MA packs have caught up with the air control. The MA pack containing 10:80:10 (O2:CO2:N2) is significantly lower than all other packs. (A–D) Each point represents the mean +/− SEM of 3 samples.

allowing extended retail. As a first step a QIM for whole N. norvegicus lobster was developed and its reliability to assess shelf life was demonstrated. The newly developed QIM mainly monitored the progress of melanosis and also the extent of off-odors and the product is rejected at 6.5 days. In fact, odor became the key parameter when we assessed product that was treated with an anti-melanotic. This is in agreement with a study by Gómez-Guillén et al. (2007). This is also in accordance with data from fish following procedures such as filleting or cooking, which preclude the assessment of the overall appearance. In such instances, the odor of the fish has also proved to be a reliable parameter for sensory assessment and rejection point determination (Lehmann and Aubourg, 2008). To identify a suitable gas mix for the MA-packaging of whole N. norvegicus we have employed a triangular un-biased approach, which gave the opportunity to test a large number of potential gas mixes. During preliminary trials, a storage regime with two storage temperatures was chosen. In this way, we identified a number of gas mixes that resulted in both reduced bacterial numbers and low QIM scores. All gas mixes identified to have a beneficial effect on shelf life were high in CO2 content. This is no surprise since a general positive effect of high carbon dioxide concentrations on the shelf life of chilled seafood is reported in the literature (Reddy et al., 1995;

Emborg et al., 2002; Ruiz-Capillas et al., 2003; Hovda et al., 2007). However, the beneficial effects of most gas mixes were only apparent at storage temperatures of 1 °C. When the storage temperature was increased from 1 °C to 6 °C the majority of gas mixes tested lost their bacteriostatic properties by day 5, and by day 8 the product was spoiled and rejected as shown by increased QIM scores and higher total bacteria counts. However, the gas mix of 10% O2, 80% of CO2 and 10% N2 still resulted in acceptable QIM scores, and exhibited low TBCs also. This gas mix (the OGM) was further tested in a series of more thorough assessments, and its full potential on extending the shelf life on whole N. norvegicus was evaluated. Since the ultimate goal of this study was to identify a gas mix for retail purposes a storage regime was chosen that simulates the processing, transportation, distribution, shelf display and household refrigerated storage conditions used in the industry. Young's Seafood Ltd (Grimsby, UK) provided this information with consideration to food safety and quality. The samples were stored on ice post harvest for 24 h, then treated and packed in a cold room and stored at 1 °C for 3 days. The packs were then transferred to 4 °C storage for 3 days and finally stored at 6 °C for the remainder of the time. Guidelines describing similar schemes are issued by the international HACCP Alliance or the Food Standards Agency in the UK (FSA) and are used

8 7

n=3x5

6 5 4 3 2

9

C

8 7

n=3x5

6 5 4 3 2 1 F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

luminous bacteria count [LOG cfu g-1]

F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

1

9

B

8

n=3x5

7 6 5 4 3 2 1

F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

A

9

D

8 7

n=3x5

6 5 4 3 2 1 F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

total bacteria count [LOG cfu g-1]

9

H2S-producing bacteria count [LOG cfu g-1]

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

Pseudomonas sp. count [LOG cfu g-1]

374

E

QIM [odour]

3

n=3x3

2

1

F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

0

Fig. 4. Bacteria counts and QIM of whole N. norvegicus lobster stored in MA or air at 1 °C. (A) shows the total bacteria count of MA-packed N. norvegicus lobster (black circles) and N. norvegicus lobster stored in air (white circles). Total counts in MA packs remain low while in the air control bacterial numbers rise continuously. (B) shows the corresponding H2S-producing bacteria count. (C) shows the bacteria count of luminous bacteria and (D) shows the Pseudomonas counts in the same samples. Pseudomonas species are completely absent from the MA packs. For (A–D) each point is the mean +/− SEM of 3 × 5 samples. (E) shows the MA-adapted QIM score for the same samples. Each point is the mean +/− SEM of 3 × 3 samples.

at various seafood companies and independent quality laboratories. For comparison, a second lot of packs were stored at a constant temperature of 1 °C, representing ‘best practice’ control. When stored at 1°C the OGM-MAP successfully doubled the shelf life of whole N. norvegicus lobster from 6.5days to 13days. Parkin and Wells (1982) reported similar results in rockfish (Sebastes species) using a comparable gas mix. In shrimp, a 100% CO2 atmosphere extended the

shelf life in much the same way (Matches and Layrisse, 1985). No strong off-odors were found in the N. norvegicus OGM-MA packs during the entire storage period. Nevertheless, by day 3 the most favorable odors described as ‘fresh, marine, hay-like described odors’ were lost. Such loss is well known in white fish and cannot be prevented as it is caused by autolytic changes within the meat (Huss, 1995). If stored at 1 °C, bacteria levels in the N. norvegicus OGM packs were generally low and

total bacteria count [LOG cfu g-1]

9

A

8 7

n=3x5

6 5 4 3 2

9

7

n=3x5

6 5 4 3 2 1

B n=3x5

7 6 5 4 3 2 1

F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14 9

D

8 7

n=3x5

6 5 4 3 2 1 F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

8

Pseudomonas sp. count [LOG cfu g-1]

C

Fr Pr esh ep a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 ay D 9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

luminous bacteria count [LOG cfu g-1]

9

375

8

F Pr res ep h a D ck ay D 1 ay D 2 ay D 3 ay D 4 ay D 5 ay D 6 ay D 7 ay D 8 a D y9 ay D 10 ay D 11 ay D 12 ay D 13 ay 14

1

H2S-producing bacteria count [LOG cfu g-1]

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

E 3

QIM [odour]

n=3x3

2

1

F Pr resh ep a Da ck y Da 1 y Da 2 y Da 3 y Da 4 y Da 5 y Da 6 y Da 7 y Da 8 Da y 9 y Da 10 y Da 11 y Da 12 y Da 13 y 14

0

Fig. 5. Bacteria counts and QIM of whole N. norvegicus lobster stored in MA or air at 1 °C for 3 days followed by 4 °C for 2 days and 6 °C for the remainder of the time. (A) shows the total bacteria count of MA-packed N. norvegicus lobster (black circles) and N. norvegicus lobster stored in air (white circles). Total counts in MA packs remain low until the temperature is stepped up to 6 °C while in the air control bacterial numbers rise continuously especially when the storage temperature is increased at 6 °C. (B) shows the corresponding H2S-producing bacteria count. (C) shows the bacteria count of luminous bacteria and (D) shows the Pseudomonas counts in the same samples. Pseudomonas species are completely absent from the MA packs. For (A–D) each point is the mean +/− SEM of 3 × 5 samples. (E) shows the MA-adapted QIM score for the same samples. Each point is the mean +/− SEM of 3 × 3 samples.

increased only slowly. In comparison, the air control was rejected on day 7 due to off-odors and very high bacterial numbers in accordance with a 6.5-day shelf life. OGM-MA packs stored according to the 3-step temperature regime performed differently to the 1 °C stored packs. For the first 5 days of storage (2 days at 1 °C followed by 3 days at 4 °C) the bacterial numbers and QIM results were comparable and no significant differences were detected. However, once the temperature was stepped up to 6 °C a

significant change was observed. Within 2 days of storage at a higher temperature (on day 7), the OGM packs contained significantly higher total bacterial counts and by day 8 the packs were rejected. This sudden change can be explained by the fact that bacterial growth rates are affected by temperature and even though the differences of 3 °C (from 1 °C to 4 °C), and 2 °C (from 4 °C to 6 °C) appear marginal, effects can be extensive. For example, an increase of 4 °C (from 4 °C to 8 °C) reduced the shelf life of cod by more than

376

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

Fig. 6. Neighborhood-joining trees of bacterial isolates found in fresh N. norvegicus tail meat and after 8 days of OGM-MA or air storage at 1 °C. The bacterial type strains are: Vibrio fischeri (ATCC 7744; accession number AY341436), V. logei (ATCC 29985; EU221273), V. splendidus (ATCC 33125; AF413024), V. wodanis (NVI 88/441T; NR_028881), Photobacterium phosphoreum (ATCC 11040; D25310), P. profundum (DSJ4; D21226), Shewanella putrefaciens (R1418; AB208055), Pseudoalteromonas haloplanktis (ATCC 14393; 67024) and Psychrobacter nivimaris (88/2–7; NR_028948). P. nivimaris was used as an out-group. The scale bars show the average number of nucleotide changes per sequence position. The clade affiliation of any unknown bacterial isolate was inferred using its relative position in relation to internal tree branches and clustering with any given type strains. Major clades are highlighted using colored boxes. The proportion of bacterial clades within a given treatment group is indicated in brackets. Note the absence of Shewanella putrefaciens isolates in both air and MA packaging (MAP) and the almost complete dominance (94%) of P. phosphoreum-like bacteria in the MA packs after 8 days at 1 °C. Neighbor-joining bootstrap values are given for the nodes separating the relevant major clades.

50% (Reddy and Solomon, 1999). The loss of shelf life extension at an elevated storage temperature observed in this study will be in part due to the known loss of the bacteriostatic effect of CO2 above 5 °C (Stammen et al., 1990; Church and Parsons, 1995; Sivertsvik and Jeksrud, 2002). CO2 has significant effects on microbial growth (Dixon and Kell, 1989) and its efficiency depends on the amount of gas solubilized into the tissue (Sivertsvik and Jeksrud, 2002). This solubility is highly temperature-dependent, and it appears that in the present case at 4 °C enough CO2 is absorbed and retained within the tissues of N. norvegicus. However, when the storage temperature is increased to 6 °C the amount of dissolved gas is reduced so that the bacteria can grow again. Many of the known SSOs such as S. putrefaciens, Vibrio species and Photobacterium species produce H2S, which causes rotten egg-like, putrid smells and sensory rejection of fish (Gram et al., 1987). We were able to analyze broadly the effects of the OGM on the bacteria microflora found in the meat of N. norvegicus during long-term storage. In the OGM-MA packs, no Pseudomonas species were detectable, which is in accordance with data from comparable MA packs of freshwater crayfish Pacifastacus leniusculus during cold storage (Wang and Brown, 1983). A large fraction of the total bacterial count in N. norvegicus is composed of H2S-producing and luminous bacteria. However, although H2S-producing and luminous bacteria respond to temperature increases during storage, the MA packs always exhibit lower counts than the air control. In all cases the OGM-MA reduced the amount of these bacteria. A more detailed assessment of the microflora using molecular tools allowed a better interpretation of the results obtained with agar plates.

The microflora of fresh samples exhibits a diverse microflora that contains various bacterial species. However, when stored for 8 days in air or MA, the bacterial diversity is drastically reduced. Photobacterium species dominate the microflora of OGM-MA packs. This is not a new finding, since P. phosphoreum is known to withstand high CO2 concentrations. Dalgaard and colleagues, for example, demonstrated that P. phosphoreum, as opposed to S. putrefaciens, is the SSO in cod (Gadus morhua) fillets (Dalgaard et al., 1993; Dalgaard, 1995). It seems that the same is true for N. norvegicus, when packed in a high CO2 atmosphere. The dominance of P. phosphoreum is less pronounced but still significant in the air control, where Vibrio species co-occur. The notorious SSO of white fish – S. putrefaciens – was not found in either the OGM-MA packs or the air control. The importance of P. phosphoreum and the absence of S. putrefaciens during spoilage of N. norvegicus have been reported elsewhere (Gornik et al., 2011), and the dominance of both P. phosphoreum and Vibrio species is reflected in the high numbers of H2S-producing and luminous bacteria found on MIA plates. 5. Conclusion We conclude that the shelf life of N. norvegicus can be extended from 6.5 to 13 days when packaged in MA with 10 parts O2, 80 parts CO2 and 10 parts N2 (10:80:10) formulation, and when storage temperatures do not exceed 1 °C. When MA packs are stored at higher temperatures (N1 °C) the shelf life gain is less pronounced and indeed becomes marginal above 5 °C. Main SSO in MA packs of N. norvegicus was shown to be P. phosphoreum in contrast to the typical SSO of fish caught

S.G. Gornik et al. / International Journal of Food Microbiology 167 (2013) 369–377

in temperate waters, namely S. putrefaciens, which does not play a significant role in spoilage of MA-packaged N. norvegicus. Despite the reported 13-day shelf life, we recommend MAP storage for N. norvegicus not to exceed 10days at 1 °C. This follows the guideline of the Advisory Committee on the Microbiological Safety of Food (ACMSF) in the UK, which since 1994 has imposed a shelf life limit of 10 days for MA packs, where the temperature is the only additional controlling factor.

Acknowledgments This work was supported by a grant from the EU Financial Instrument for Fisheries Guidance (FIFG) Scheme through the Scottish Executive and by Young's Seafood Ltd. The authors thank the skipper and staff of Research Vessel ‘Aplysia’ at the University Marine Biological Station Millport for help with the animal collection. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijfoodmicro.2013.10.002.

References Albalat, A., Gornik, S.G., Mullen, W., Crozier, A., Atkinson, R.J.A., Coombs, G.H., Neil, D.M., 2011. Quality changes in chilled Norway lobster (Nephrops norvegicus) tail meat and the effects of delayed icing. Int. J. Food Sci. Technol. 46 (7), 1413–1421. Barbosa, A., Vaz-Pires, P., 2004. Quality index method (QIM): development of a sensorial scheme for common octopus (Octopus vulgaris). Food Control 15 (3), 161–168. Bartolo, I., Birk, E.O., 2002. Some factors affecting Norway lobster (Nephrops norvegicus) cuticle polyphenol oxidase activity and blackspot development. Int. J. Food Sci. Technol. 33 (3), 329–336. Bremner, H.A., 1985. A convenient easy to use system for estimating the quality of chilled seafood. Fish Process. Bull. 7 (1), 59–70. Cardenas Bonilla, A., Sveinsdottir, K., Martinsdottir, E., 2007. Development of Quality Index Method (QIM) scheme for fresh cod (Gadus morhua) fillets and application in shelf life study. Food Control 18 (4), 352–358. Church, I.J., Parsons, A.L., 1995. Modified atmosphere packaging technology: a review. J. Sci. Food Agric. 67 (2), 143–152. Dalgaard, P., 1995. Qualitative and quantitative characterization of spoilage bacteria from packed fish. Int. J. Food Microbiol. 26 (3), 319–333. Dalgaard, P., Gram, L., Huss, H.H., 1993. Spoilage and shelf-life of cod fillets packed in vacuum or modified atmospheres. Int. J. Food Microbiol. 19 (4), 283–294. Dixon, N.M., Kell, D.B., 1989. The inhibition by CO2 of the growth and metabolism of micro-organisms. J. Appl. Microbiol. 67 (2), 109–136. Emborg, J., Laursen, B.G., Rathjen, T., 2002. Microbial spoilage and formation of biogenic amines in fresh and thawed modified atmosphere-packed salmon (Salmo salar) at 2 °C. J. Appl. Microbiol. 92 (4), 790–799.

377

Giménez, B., Martínez-Alvarez, Ó., Montero, P., Gómez-Guillén, M.D.C., 2010. Characterization of phenoloxidase activity of carapace and viscera from cephalothorax of Norway lobster (Nephrops norvegicus). LWT Food Sci. Technol. 43 (8), 1240–1245. Gómez-Guillén, M.C., López-Caballero, M.E., Martínez-Álvarez, Ó., Montero, P., 2007. Sensory analyses of Norway lobster treated with different antimelanosis agents. J. Sens. Stud. 22 (6), 609–622. Gornik, S.G., Albalat, A., Macpherson, H., Birkbeck, H., Neil, D.M., 2011. The effect of temperature on the bacterial load and microbial composition in Norway lobster (Nephrops norvegicus) tail meat during storage. J. Appl. Microbiol. 111 (3), 582–592. Gram, L., Trolle, G., Huss, H.H., 1987. Detection of specific spoilage bacteria from fish stored at low (0 °C) and high (20 °C) temperatures. Int. J. Food Microbiol. 4 (1), 65–72. Hovda, M.B., Sivertsvik, M., Lunestad, B.T., Lorentzen, G., Rosnes, J.T., 2007. Characterisation of the dominant bacterial population in modified atmosphere packaged farmed halibut (Hippoglossus hippoglossus) based on 16S rDNA-DGGE. Food Microbiol. 24 (4), 362–371. Huidobro, A., Pastor, A., Tejada, M., 2000. Quality index method developed for raw gilthead seabream (Sparus aurata). J. Food Sci. 65 (7), 1202–1205. Huss, H.H., 1995. Quality and Quality Changes in Fresh Fish. FAO Fisheries Technical Paper, 348. Hyldig, G., Bremner, A., Martinsdóttir, E., 2007. Quality index methods. In: Nollet, L.M.L. (Ed.), Handbook of Meat, Poultry & Seafood Quality. Blackwell Publishing, Oxford, pp. 529–549. Lehmann, I., Aubourg, S.P., 2008. Effect of previous gutting on rancidity development in horse mackerel (Trachurus trachurus) during frozen storage at −20 °C. Int. J. Food Sci. Technol. 43 (2), 270–275. Martínez-Álvarez, Ó., López-Caballero, M.E., Montero, P., Gómez-Guillén, M.C., 2007. Spraying of 4-hexylresorcinol based formulations to prevent enzymatic browning in Norway lobsters (Nephrops norvegicus) during chilled storage. Food Chem. 100 (1), 147–155. Martínez-Álvarez, Ó., Gómez-Guillén, M.D.C., Montero, P., 2008. Chemical and microbial quality indexes of Norwegian lobsters (Nephrops norvegicus) dusted with sulphites. Int. J. Food Sci. Technol. 43 (6), 1099–1110. Matches, R.J., Layrisse, E.M., 1985. Controlled atmosphere storage of Spotted shrimp (Pandalus platyceros). Journal of Food Production 48 (8), 709–711. Neil, D.M., 2012. Ensuring crustacean product quality in the post-harvest phase. J. Invertebr. Pathol. 110 (2), 267–275. Parkin, K.L., Wells, M., 1982. Modified atmosphere storage of rockfish fillets. J. Food Sci. 47 (1), 181–184. Reddy, N.R., Solomon, H.M., 1999. Comparison of margin of safety between sensory spoilage and onset of Clostridium botulinum toxin development during storage of modified atmosphere (MA)-packaged fresh marine cod fillets with MA-packaged aquacultured fish fillets. J. Food Saf. 19 (3), 171–183. Reddy, N.R., Villanueva, M., Kautter, D.A., 1995. Shelf life of modified-atmospherepackaged fresh tilapia fillets stored under refrigeration and temperature-abuse conditions. J. Food Prot. 58 (8), 908–914. Ruiz-Capillas, C., Morales, J., Moral, A., 2003. Preservation of bulk-stored Norway lobster at 1 °C in controlled and modified atmospheres. Eur. Food Res. Technol. 217 (6), 466–470. Sivertsvik, M., Jeksrud, W.K., 2002. A review of modified atmosphere packaging of fish and fishery products — significance of microbial growth, activities and safety. Int. J. Food Sci. Technol. 37 (2), 107–127. Stammen, K., Gerdes, D., Caporaso, F., 1990. Modified atmosphere packaging of seafood. Crit. Rev. Food Sci. Nutr. 29 (5), 301–331. Wang, M.Y., Brown, W.D., 1983. Effects of elevated CO2 atmosphere on storage of freshwater crayfish (Pacifastacus leniusculus). J. Food Sci. 48 (1), 158–162.