Potential fish decontamination treatments

Currently, there is a lack of evidence on the effectiveness of smoky water vapour applied without significant heat on the numbers of L. monocytogenes associated with smoked fish. However, it has been proposed that the process of cold smoking could be altered to include a steaming step (Dimitradou et al. ,2008). In fish steamed in a mixture of liquid smoke and water, the numbers of naturally present total aerobic mesophiles (TAMC) reduced from an initial 5.9x105cfu/g to 25 cfu/g. After prolonged storage for 91 days at 4oC, if the fish had not been previously dried, the TAMC was still 25 cfu/g. However, there was no L. monocytogenes contamination of the fish by either artificial or environmental routes therefore, although the treatment has benefit for increased shelf life, there is a lack of evidence as to the effectiveness of steaming in reducing L. monocytogenes numbers.

Shi et al. (2015) undertook experimental work that attempted to determine how heat applied to cold smoked fish inactivated L. monocytogenes. The work was undertaken from the viewpoint of the development of a pasteurisation-style process for fish. Cold-smoked salmon was inoculated with a cocktail of three strains of L. monocytogenes and one strain of L. innocua. Inactivation curves were obtained for 58, 60, 62, 64, and 66oC. Inactivation was typically log linear. D-values (the time taken for the population to decimate by one order of magnitude) for both L. monocytogenes and L. innocua were 0.3 to 14.1 min at 66 to 58oC, with a z-value (the number of degrees the temperature had to be increased to achieve a tenfold reduction in the D-value) of 5.2 to 6.5oC. The non-pathogenic Listeria innocua had comparable D- and z-values to the three strains of L. monocytogenes and thus can be used for validation of pasteurisation processes to control L. monocytogenes in cold-smoked salmon. Shi et al. (2015) briefly discussed the consequences of a thermal process that could cause a 6-log reduction in L. monocytogenes. Such a process would require a 7 min full exposure to 64oC or a 4 min exposure at 66oC, which are times and temperatures similar to cooking. As the authors say, “Pasteurisation under these conditions is likely to have a significant impact on the sensory characteristics that cold-smoked salmon is prized for”.

The effect of freezing stress on L. monocytogenes has been studied (Yoon et al. 2004). Cold smoked salmon fillets inoculated with L. monocytogenes were frozen at -20oC for five days and subsequently stored at 4 and 10oC for up to 60 days. The freezing treatment increased the lag phase before L. monocytogenes growth by 10-15 days when stored at 4oC, and four days when stored at 10oC. The numbers of freeze-stressed L. monocytogenes stored at 4oC never reached those of the non-frozen controls, even after 60 days. However, when refrigeration was at 10oC and after as little as 16 days, L. monocytogenes numbers increased to more than 7 log10 cfu/g fish which was similar to numbers achieved by the unstressed controls.

Guyer and Jemmi (1991) undertook three separate trials during which they observed the growth of inoculated L. monocytogenes on salmon fillets during their processing and storage. There were no significant differences between the growth of a reference strain and a salmon fillet-derived strain of L. monocytogenes used for the studies. A general conclusion of the work was that freezing of the finished product, followed by thawing and refrigerated storage had no significant effect on the numbers of inoculated L. monocytogenes compared with an unfrozen inoculated control.

Kang et al. (2012) have more recently reported on the effect of freezing on L. monocytogenes lag times on fish stored at 7oC. Salmon fillets were inoculated with a lab-cultured four strain cocktail L. monocytogenes after a freeze-thaw cycle and cellular growth compared with inoculated fillets that had been chilled only. An important distinction between the Guyer and Jemmi (1991) work and the Kang et al (2012) study is when the fish were inoculated. For Guyer and Jemmi (1991), it was prior to curing; for Kang et al (2012) if was after a freeze and subsequent thaw. The Guyer and Jemmi (1991) work is a model for contaminated raw fish arriving at a processor. For the Kang (2012) study the L. monocytogenes cells were not stressed by the process or injured by freezing. Therefore, the Kang study models the contamination of L. monocytogenes after frozen storage. In the UK it is usual to vacuum package salmon prior to frozen storage, and so unless there was damage to packaging, or a repacking of finished product to meet customer supply criteria, the scenario modelled assessed by Kang would be unlikely to occur in the UK. Some L. monocytogenes are able to use one or a variety of non-gaseous terminal electron acceptors in place of oxygen and so vacuum packing is not a reliable intervention for L. monocytogenes.

In contrast to an observation of no significant effect reported by Guyer and Jemmi (1991), Kang et al. (2012) reported that frozen fillets supported statistically significant increased L. monocytogenes growth rates over 30 days imperfect storage at 7oC. Thus, the work of Kang et al (2012) highlights a potential hazard if freeze-thawed smoked salmon become contaminated with L. monocytogenes and are imperfectly refrigerated.

Super-chilling involves reducing the temperature of fish uniformly to a point slightly below that obtained in melting ice and has been used to extend the shelf life of the fish. Beaufort et al. (2009) found that super chilling of smoked salmon to -2oC for 14 days reduced the prevalence of L. monocytogenes to 9.0% compared with 39.0% for storage above 0oC. However super chilled fish had a minor effect on the organoleptic properties of the fish and L. monocytogenes could still exceed the EU limits, even if storage after a super chilling treatment was at 4oC.

In summary, the treatment of smoked fish by temperature reduction to below 0oC appears to have a fairly limited impact on L. monocytogenes proliferation during subsequent storage at higher temperatures: a below zero cold treatment followed by refrigerated storage does not eradicate  L. monocytogenes from smoked fish. The best-case scenario is that if the subsequent storage temperatures are low enough, the below zero temperature treatment can extend the amount of time before  L. monocytogenes commences exponential growth. Freezing and super cooling as treatments followed by refrigerated storage are not reliable interventions for  L. monocytogenes control. Keeping fish frozen with thawing only at the point of use of course prevents  L. monocytogenes growth for the frozen stored period.

Food irradiation is a processing technique which exposes food to electron beams, X-rays, gamma rays or other electromagnetic wavelengths. The radiation exposure causes reductions in numbers of microorganisms, with high exposures effectively sterilising food. In contrast to pasteurisation, cooking or other forms of heat treatment, irradiated food tends not to change colour or have an altered texture, although there can be minor chemical changes to foods subjected to the process. A great deal of research spanning several decades funded by the WHO, UN-FAO and the USDA has shown that irradiation of food is generally safe and an effective way to kill bacteria and preserve food.

Irradiated foods in the EU are subject to some strict controls applied at the individual member state level. In the UK, the national and devolved Food Standards Agencies licence premises to use irradiation subject to approval by the European Commission. At the time of writing, there is only one food processing premises in the UK with a licence to irradiate food and its authorisation is restricted solely to herbs and spices. The Food Irradiation Regulations (2009) oblige food processors to clearly label food exposed to ionising radiation and restrict irradiation only to a number of food classes.

Fish and shellfish are an allowed food group which can be exposed to up to 3 kGy provided:

• there is a reasonable technological need
• the food presents no health hazard, and the radiation treatment is not used as a substitute for hygiene and health practices or for good manufacturing or agricultural practices
• the irradiation benefits consumers

In addition, other restrictions apply to ensure that food irradiation may only be used to:

• reduce the incidence of food-borne disease by destroying pathogenic organisms
• reduce spoilage of foodstuffs by retarding or arresting decay processes and destroying spoilage organisms
• reduce loss of foodstuffs by premature ripening, germination or sprouting
• rid foodstuffs of organisms harmful to plant or plant products

E-Beam irradiation

Electron beam (e-beam) irradiation uses electrons to irradiate food and reduce the numbers of microorganisms associated with it. The process targets microorganisms’ nucleic acids and is becoming popular in the USA because it is a cold process which does not significantly alter the structure or flavour of a number of foods. Informally in the USA, the process is referred to as ‘cold pasteurisation’ and can be used to inactivate L. monocytogenesin cold smoked fish (Medina et al., 2009). An initial L. monocytogenes inoculum of 9 Log cfu/g on cold smoked salmon was decreased by 7 log units after exposure to a dose 4 kGy of e-beam radiation. A dose of 1 kGy produced a 2-log inactivation (Medina et al. 2009). However, in contrast to many foods, E-beam doses of 2 to 4 kGy generate obvious ‘off odours’ for cold smoked salmon.

Skowron et al. (2018) assessed gamma radiation and high energy electron beam doses on the inactivation of antibiotic-susceptible and antibiotic-resistant L. monocytogenes strains inoculated onto the surface of raw salmon fillets and stored at different temperatures (-20oC, 4oC and 25oC). Consideration of AMR is becoming increasingly important as the available choices of antimicrobial chemotherapies to treat infections continues to narrow. The approach used by the Skowron study is unorthodox, which is a potential criticism of the work. 

In essence, strains were exposed to increasing sub-MIC concentrations of antibiotics and an AMR status conferred by selection by the exposure rather than using naturally resistant strains containing AMR genes in plasmids or transposons. A key finding of the study was that the lethal doses for both radiation methods were higher for the antibiotic-resistant strains compared with sensitive strains. Gamma irradiation was more effective than electron beam exposure for extending the shelf-life of salmon fillets. PFGE revealed that the repair to DNA damaged by exposure to the radiation occurred faster in the antibiotic-resistant L. monocytogenes strains. Irradiated AMR strains grew more rapidly than non-AMR irradiated strains after exposure to the radiation. Currently within the EU, irradiation of foods is confined to a narrow range of product, which includes herbs, but not fish.

X-Ray irradiation

The effectiveness of X-ray irradiation on reducing L. monocytogenes numbers in ready-to-eat vacuum packed smoked mullet has been assessed (Robertson et al. 2006). It was found that a dose of 2 kGy was required to eliminate an initial level of 104 cfu/g. No change of sensory quality was detected in the product.

More recently in Eygpt, Badr (2012) reported the effect of exposing cold smoked salmon inoculated with lab-cultured L. monocytogenes to x-rays. The work was broadly similar to the previous studies of Robertson et al (2006), and the findings from both studies broadly agree. In essence, Badr (2012) observed around a six-log reduction in the L. monocytogenes population on the fish when the x-ray dose was 3 kGy. Organoleptic assessment of un-inoculated fish exposed to the same dose by an untrained panel of ten tasters did not detect a significant change to the texture, colour, odour, or flavour of the fish.

In the USA, Mahmoud (2012) added to the growing body of evidence that irradiation can delay the growth of L. monocytogenes. An initial laboratory-cultured inoculation of L. monocytogenes was applied at a concentration of 3.7 log cfu/g fish. After exposure to a 1.0 kGy dose of x-ray the L. monocytogenes population was reduced to below the limit of detection of the quantitative test method. However, the L. monocytogenes was apparently only sub-lethally injured, recovered and counts gradually increased during subsequent storage. Treatment with 2.0 kGy X-ray kept the L. monocytogenes population under the limit of detection for 35 days. For the Mahmoud (2012) study, no organoleptic testing was undertaken to determine the consequences of the irradiation.

Although there are limited data, a brief summary of these studies are that exposure of cold smoked salmon to doses of 1 kGy of x-rays decreases natural bacterial populations and consequently increases the shelf-life by about 20 days with only slight off-odours detectable after 35 days storage at 5oC. A dose of 2-3kGy has a significant impact on high numbers of L. monocytogenes. In addition, X-ray appeared a suitable treatment for smoked fish because it did not leave any significant organoleptic evidence of the intervention. Whether smoked fish could be irradiated in the EU is questionable because potentially, L. monocytogenes could be a health hazard to vulnerable groups and the UK irradiated food regulations (2009) state that any treated food should presents no health hazard prior to irradiation. Finally, it is considered unlikely any UK fish smoker would irradiate fish currently. In addition to licencing requirements, in the EU, foods treated in such a manner need to be clearly labelled as such and the average consumer lacks technical knowledge of the technology, which limits uptake (Eustice and Bruhn, 2010).

High pressure processing (HPP) is also known as High Hydrostatic Pressure (HHP) and is a process that has been shown to be effective in reducing numbers and delaying the regrowth of L. monocytogenes associated with smoked fish (Tocmo et al. 2014). Using a four-strain cocktail of L. monocytogenes, an initial study showed there was a correlation between growth lag for L. monocytogenes and pressure applied (Lakshman and Dalgaard 2004). Samples treated with 250 MPa had a 17-day lag period before L. monocytogenes growth commenced compared with one day for untreated controls. However, and texture and colour differences were seen at pressure applications of 200 MPa of pressure or higher.

However, it became clear after subsequent work that pressure, salt and phenol act synergistically to inhibit L. monocytogenes. No bactericidal effect was achieved when dolphinfish which had been smoked under mild conditions (1.97% salt and 42 ppm phenol) was exposed to a high-pressure treatment of 300 MPa at 20oC for 15 min. However, under more severe salting and smoking conditions (2.93% salt and 82 ppm phenol), pressurization kept L. monocytogenes counts under the detection limit throughout 100 days of storage. Both high pressure (Lakshman and Dalgaard, 2004) and increased phenolic compound concentrations (Vitt et al. 2001) have previously been reported as causing unacceptable organoleptic changes to the product.

Temperature and pH also influence the effectiveness of pressure. A French study investigated the effects of high-pressure processing at 100, 150, and 200 MPa combined with sub-zero temperatures of -10oC, -14oC, and -18oC at pH 7.0 and pH 4.5 on L. monocytogenes present on salmon fillets (Ritz et al. 2008). Perhaps not surprisingly, the study showed that the most effective high-pressure treatment for L. monocytogenes inactivation was when extremes for high pressure and low pH and temperature were used. Under these conditions, L. monocytogenes numbers were lowered more than six logs rather than extending the lag phase before growth (Ritz et al. 2008).

However, modifications of the physical properties of the fish flesh were a consequence of the treatment. In particular, the fish needed to be heated before the pressure was released to prevent freezing. Consequently, the researchers observed a pronounced lightening of the pinkish colour of the flesh as well as an increased toughness of the meat which they believed may be acceptable to consumers on the grounds that they were indicators of improved food safety.

High pressure alone (450 MPa) over a period of 10 minutes at 12oC was shown by Medina et al. (2009) to produce an initial three log decrease in L. monocytogenes artificially inoculated onto cold smoked salmon by dipping into a solution containing 3 x106 cfu L. monocytogenes per ml. However, the L. monocytogenes recovered and after 35 days at 5oC growth of at least a log was observed during storage at 5oC. If the storage temperature was 5oC - 8oC only 21 days were required for a one log growth. However, Medina et al. (2009) did note that there was no odour given off as a result of the treatment although slight colour changes were observed. A potential criticism of the study was that only a single strain of L. monocytogenes isolated from chicken was used.

The potential of HPP was also investigated by Gudbjornsdottir et al. (2010) who found that high pressure treatment (700-900 MPa) was of potential value for the control of Listeria associated with fish. In the Gudbjornsdottir study, L. innocua was reduced from 4.5 x 103 cfu/g to undetectable numbers. Although no substantial change in the colour of the fish flesh was observed except a minor lightening of the product, the microstructure of the cold smoked salmon was most detrimentally affected at the highest pressure 900 MPa with a treatment time of 60s (Gudbjornsdottir et al. 2010). In keeping with the other studies (Ritz et al. 2008; Medina et al. 2009) the high-pressure treatment caused changes in the fish flesh structure resulting in hardening of the fish flesh severe enough to be perceived by consumers (Gudbjornsdottir et al. 2010).

Mengden et al. (2015) reported another assessment of HPP as control treatment for L. monocytogenes in cold smoked rainbow trout fillets and fresh European catfish fillets. Treated fish inoculated with L. monocytogenes or E. coli were stored under refrigerated conditions for 41 or 7 days, respectively. The assessed treatments were mincing of the fish before exposure to 200, 400, or 600 MPa for 1 or 5 min at room temperature. HPP significantly reduced L. monocytogenes by >6 log10 cfu/g in both fish products, but subsequent growth was detected. Reductions of E. coli were also significant and of the order of >6 log10 cfu/g but during refrigerated storage, no growth was evident in samples. Spoilage microbiota were significantly reduced in the catfish fillets whereas the counts in trout fillets were only 1-2 log10 cfu/g throughout storage. In contrast to the nearly unaffected trout fillets, HPP-treatment catfish fillets appeared to be paler with the appearance of a cooked product. Taking the results of the favourable sensory analyses into account, HPP was assessed as suitable for the decontamination of mild smoked rainbow trout fillets at the higher end of the assessed treatments (e. g. using 600 MPa for five minutes).

Ekonomou et al. (2020) also evaluated various combinations of decontamination treatments of high hydrostatic pressure (HPP; 200 MPa for 15 min), liquid smoke (0.50%, v/v) and freezing (−80 °C, overnight) on the numbers of L. monocytogenes in BHI broth, raw and smoked trout. The listeriacidal effect of commercially available liquid smoke solutions (L9 and G6), HPP and their combinations were evaluated against three L. monocytogenes strains individually. The most resistant strain (10403S) was subjected to a combined treatment. A synergistic effect of liquid smoke and HPP was observed and was further enhanced by freezing overnight prior to HHP. The effect of HPP and liquid smoke, prior to freezing was highest in BHI compared to raw and smoked trout, which may be evidence that further supports the findings of Chen et al. (2020). The three treatments applied in combination provided either a 5.48 (smoked trout) or 1.93 log cfu/g reduction (raw trout) were used. There are some considerations that were not made by the authors. The first is that liquid smoke composition is variable between batches. There was insufficient replication with different smoke batches to determine if the study findings related to just the batch of liquid smoke used.

The discussions within the Ekonomou paper are interesting. The situation for using high pressure is less than clear in the EU. There have been discussions within the EU on the legality of high pressure, but the most recent consolidations of documents (such as EC 852/2004) do not make any specific mention of HHP. Further investigations outside of the the Ekonomou publication have revealed there is HHP treatment of foods in France, but the products are exported outside of the EU. There is a 2010 academic thesis that considered the legality of HPP. The ultimate conclusion was that it was a matter that can be decided by individual EU member states on a case-by-case basis.

The Irish government apparently concurred and have provided guidance for industry on the requirements for food processing using HPP in Ireland in its capacity as an EU member state. Thus, these papers collectively describe a treatment that may well be legal in the EU if it is permitted by the national government, and which can act as a critical control point for cold smoked fish. Before high pressure processing could be properly assessed as a practical intervention, formal taste testing would need to be undertaken to determine if the reported physical changes to the product, such as changed colour were acceptable to consumers. In addition, the food may have to be labelled as pressure treated, to comply with regulation EC 1169/2011.

Jasim and colleagues (2021) applied high pressure (375 MPa for 20 min) to fish. The treatment was determined as optimal to destroy L. monocytogenes that had been artificially inoculated onto the fish. The paper described the associated changes to amino acids and lipids that occurred as a consequence of the HPP and so was not strictly focussed on L. monocytogenes. The HPP treatment significantly increased the total amino acids from 5.89 to 6.92 mmol/g, although there was no significant change in the free amino acids. The HPP treatment did not influence the fatty acid profiles.

Pulsed light involves the use of very high intensity and short duration pulses of broad-spectrum ‘white light’ that last only a few hundred millionths of a second. In the EU, use of pulsed light would likely be subject to compliance with the consumer safeguards originally defined in regulation EC 258/97 and updated and revised as EC 2015/2283. The key issue is whether the treatment would be classed as irradiation by the regulators. Currently, whether that is the case is uncertain.

The wavelengths applied included a section of the electromagnetic spectrum normally filtered out by the Earth’s atmosphere. Since the ‘light’ contains radiation of frequencies that are not naturally found on the planet’s surface, there are few bacterial defence mechanisms that protect against it. A specific subset of wavelengths casually referred to as pulsed UV light has been specifically approved for use with food in the United States by the Food and Drug Administration.

Ozer and Demirci (2006) exposed raw (not smoked) salmon fillets inoculated with L. monocytogenesto pulses of high intensity light. The fillets were exposed at different distances from the strobe and also for different lengths of time. The authors found no differences in L. monocytogenes population reductions between the skin and muscle of salmon fillets. The reductions were however quite modest and of the order of a single log unit at a distance of 8 cm from the light strobe and a 60s exposure (Ozer and Demirci, 2006). At shorter distances (3 to 5 cm) from the strobe, high enough temperatures were recorded in the fillets to partly cook the flesh. The authors concluded a 60s treatment that was 8cm from the strobe could achieve a single log reduction without quality issues. However, at a distance of 8cm, a high intensity pulse was required. Although there is modest microbiological benefit for the use of pulsed light at that distance, the authors note that the generation of high intensity pulsed light is energy intensive and that a significant barrier to uptake is the cost of the required electricity.

Short wavelength UV-C (100-250nm) has been explored as a decontamination treatment for fish (Bernbom et al., 2011). In the EU, use of UV would likely be subject to compliance with the consumer safeguards originally defined in EC 258/97 and updated and revised as EC 2015/2283. Use of a ceiling-mounted, light source in a cold smoked salmon production plant reduced L. monocytogenes numbers on fish which were close to the light source. The reduction was time-dependant with a three-log decrease in total bacterial counts after 48 hours of exposure. The numbers of samples which tested positive for L. monocytogenes were not significantly lowered after 7 hours exposure to a UV-C lamp. In contrast, significant reductions in positive test results from fillets in close proximity to UV source were noted after 48 hours exposure. Furthermore, after the extended exposure of the fillets for 48 hours, areas greater than 5 metres away from the source or which did not have any apparent direct UV-C illumination showed a significantly reduced incidence of L. monocytogenes. The effectiveness of UV-C light was decreased by the presence of organic materials (Bernbom et al., 2011). The study acknowledged the human health implications of having a UV-C source close to workers in a production facility. This hazard could be minimized by placing plant conveyor belts inside a UV-C tunnel. Currently and subject to discussions with EHO, there are no apparent legal barriers to UV illumination. UV illumination of other foods of animal origin is widely used in other sectors such as chicken processing.

Colejo et al. (2018) also assessed the effectiveness of UV-C (254nm wavelength) and non-thermal atmospheric plasma (NTAP) treatments, applied individually or in combination, for the inactivation of a range of bacteria including L. monocytogenes. The impact of the treatments was compared when the bacteria was grown on cold smoked salmon or BHI agar. There was no specific discussion of stress responses, but the results support a number of points made by the Chen et al. (2020) paper. All microorganisms exhibited a higher resistance to UV-C and NTAP treatments on smoked salmon compared with growth media. The Colejop discussion focussed on surface roughness and porosity as the underlying reason for the difference. Exposure to an individual UV-C treatment or a highly impractical 15 min NTAP treatment reduced bacterial numbers by 1.3 logs at most, with no significant changes in the sensory quality properties of the treated products. The sequential application of UV-C and NTAP resulted in synergistic reduction in bacterial numbers. However, the combined treatments induced lipid oxidation and adverse changes in appearance and so were not suitable for commercial operations. It is well established that UV exposure turns lipids rancid in meat and fish and so is not a procedure well suited for the preservation or decontamination of foods of animal origin. The maximum reduction achieved was around 1.5 logs for a combination of UV-C and NTAP, which is of limited practical use and, legal issues aside, could only considered as part of multiple hurdles strategy. The required four-minute (Eicher et al. 2020) plasma exposure is probably impractical for adoption in a commercial processing environment.

Holck et al. (2018) used a 10-strain cocktail of L. monocytogenes, which was inoculated onto cold smoked salmon at 104 cfu/ sample and exposed to either UVC or high intensity pulsed UV light. The observed reductions in the numbers or L. monocytogenes ranged between 0.7 log cfu/ sample and 1.3 log cfu/ sample, which was similar to the reductions reported by Colejo et al. (2018). Higher reductions were observed on muscle surfaces compared with skin and scale. L. monocytogenes surviving UV treatments on smoked and raw salmon grew at the same rate as unexposed controls, although took longer to achieve statutory limits because the starting numbers were lowered by UV exposure during a storage at 4oC. No sensory difference was detected between UV-treated and untreated controls using a panel of 40 tasters immediately after UV treatment. No further assessment was made of any rancidity over the shelf life of the fish.

Although not strictly UV light, the combination of riboflavin (vitamin B2) and bluish light of 460nm wavelength was investigated by Josewin et al. (2018) on populations of L. monocytogenes inoculated onto smoked salmon. A four-strain cocktail of L. monocytogenes was used. Riboflavin solutions of concentration 25, 50 and 100 μM were topically applied to the fish surface after the L. monocytogenes was inoculated. The fish was illuminated by 460 nm LEDs using irradiance intensities of 15, 31 and 58 mW/cm2. The AM treatments were undertaken at two temperatures of 4oC and 12oC. Overall, all of the treatments showed a reduction of around one log cfu/cm2 with minor differences between different light intensities and riboflavin concentrations. There was a synergy of kill observed when both the riboflavin and illumination treatments were applied in combination. The authors believed the results of their work demonstrated a potential for 460 nm LEDs aided by riboflavin to minimise the risk of L. monocytogenes associated with smoked salmon during storage.

There is little published literature relating to the influence of natural pH (i.e. pH not manipulated by acetate, lactate or other preservative) and water activity (aW) on the survival and propagation of L. monocytogenes associated with smoked seafoods. However, a set of key papers from a Belgian research group has attempted to model the behaviour of L. monocytogenes as a function of aW, and pH based on laboratory test results from 26 clinical isolates suspected of causing foodborne illness or food isolates of L. monocytogenes (Gysemans et al. ,2007; Vermeulen et al. ,2007, and Vermeulen et al., 2009).

The growth boundary of L. monocytogenes was observed not to be restricted to a narrow transition zone, indeed “in the studied region, aw did not have a pronounced influence on the position of the growth/no growth boundary while a low concentration of acetic acid (0.2% (w/w)) and a pH decrease to 5.8 was sufficient to significantly reduce the possibility of growth” (Table 1). The findings that L. monocytogenes is tolerant to high salt concentrations and low water availability is supported by other studies in the literature (Peterson et al., 1993; Truelstrup Hansen et al. 1998; Jørgensen and Huss, 1998; Gram, 2001).

The size of the initial inoculum is a major determining factor for L. monocytogenes survival or growth in food (Vermeulen et al. 2009). Vermeulen et al (2009) showed an increasing probability of growth as the initial cell count increased, but also that pH and aW influenced the probabilities of growth too. The authors further reported that it was unlikely that a larger L. monocytogenes population had a greater chance of growth as a consequence of a higher probability in a larger population than a single stress-resistant cell multiplying. The Belgian studies concluded that it “seemed the bacteria influenced each other's growth”; which in is broad agreement with other observations regarding the importance of sessile L. monocytogenes (Tomkin 2002; Klaeboe et al., 2010).

Although sparse, there is a small amount of data describing typical pH and water activity values for smoked seafood sold in the UK (Table 2). Salt concentrations were observed to vary by product type, given shelf life and country of origin. In the UK, the salt contents of cold smoked salmon sampled at retail were recently found to range from 2.2-3.5% and had shelf lives from 10-16 days. An earlier MAFF (1991) study of ‘The microbiological status of some mail order foods’ reported salt concentrations ranging from 3.29-8.11% and shelf lives from 11-20 days. Based on the work of the Belgian researchers, the reported salt concentrations and aw typical of smoked fish in the UK, it is considered unlikely that the growth of L. monocytogenes on fish would be impacted significantly (Table 2).

Table 1: The results from screening of 26 L. monocytogenes strains to determine the aW, and pH combinations where growth did not occur.

Table is reproduced from Vermeulen et al. 2007. 

Key: 

a Culture collection identifier

b Clinical (C) or food (F) isolate

c No data was available

Table 2: Details of cold smoked fish products sold in the UK

Source: Industry data (published in Peck, Goodburn, Betts, Stringer, 2006). VP is vacuum packed; MAP is modified atmosphere packaging.

Vacuum packing

L. monocytogenes is not effectively controlled by vacuum packaging (EFSA Opinion). L. monocytogenes numbers were observed to increase by approximately one log when unsmoked fish was inoculated with L. monocytogenes, vacuum packed and stored at 4oC for one week (Rørvik, 1991). Under similar storage conditions after five weeks, populations of L. monocytogenes were shown to increase by as much as four log units (Rørvik, 1991). In combination, these findings strongly suggest that vacuum packing does not significantly inhibit the growth of L. monocytogenes in either the short or longer terms when the bacterium is using fish as a nutrient source. A potential criticism of the study was the use of an artificial inoculation of L. monocytogenes.

Conflicting observations have been reported on the implications of smoking fish prior to vacuum packing Cortesi et al., 1997; Nilsson, 1997; Beaufort et al., 2007;). Nilsson et al. (1997) found that numbers of L. monocytogenes increased from three logs to eight logs cfu/g in vacuum packed cold-smoked salmon over refrigerated storage of eight days. In contrast, Porsby et al. (2008) found that after brining and cold smoking, the numbers of L. monocytogenes decreased when the fish were vacuum-packed and stored at 5oC. There appear to be a number of common issues with reports of the fate of L. monocytogenes on fish which is vacuum-packed and stored under refrigeration. A large number of the studies (Rørvik 1991; Nilsson, 1997; Guilbaud et al. 2008; Yilmaz, 2009) use laboratory-grown L. monocytogenes strains grown in liquid broth whereas natural contamination is more likely to be from plant environment persistent sessile L. monocytogenes. In addition, artificially contaminated fish tend to be inoculated with much larger numbers of L. monocytogenes than naturally contaminated fish (Beaufort et al., 2007). Lappi et al. (2004a, 2004b) summarises the other issues as: a unified overview of L. monocytogenes growth during storage in naturally contaminated smoked fish has been difficult to interpret due to the heterogeneity of L. monocytogenes distribution (Cortesi et al., 1997) within samples (Lappi et al., 2004a) and the variable composition of smoke and smoke condensates (Sunen et al., 2003; Stołyhwo and Sikorsky, 2005).

In order to address these confounding issues, Lappi et al., (2004a) sought to closely mimic the conditions in a production scenario by observing the fate of L. monocytogenes that had naturally contaminated smoked salmon along with refrigerated storage for 28 days under vacuum. For the conditions studied, Lappi and colleagues reported that numbers of L. monocytogenes did not exceed the EU Regulation EC 2073/2005 statutory limit for ready to eat foods of 100 cfu/g fish flesh (Lappi et al., 2004a).

A later, larger French study also determined numbers of L. monocytogenes likely to be present on cold smoked salmon after vacuum packing and extended refrigerated storage (Beaufort et al. 2007). This study determined initial L. monocytogenes presence and numbers for more than one thousand naturally contaminated samples which were sourced from nine French smoking plants sporadically over a four-year period. Samples were initially tested between three and eight days after smoking and the packs were resealed. A second test was undertaken for initially positive packs after refrigerated storage for eight to 15 days at 4oC to mimic cold chain transport and the packs were again resealed. A final retest was undertaken after 8oC storage for 7 days to mimic domestic refrigeration conditions. Although initial detection prevalences in the smoked salmon ranged from 0% to 41%, more than 92% of samples contained L. monocytogenes numbers that were below 1 cfu/g fish. After the 4oC storage, there were no significant changes in L. monocytogenes prevalence or numbers. The highest numbers of L. monocytogenes observed were 7 cfu/g fish. After the 8oC storage, 17% of the contaminated products exceeded 100 cfu/g fish with the highest number observed being 2800 cfu L. monocytogenes /g fish. The key findings of the Beaufort study are that if L. monocytogenes are present on fish that are vacuum packed, even in numbers less than 1 cfu/g fish, there is the potential for growth during low temperature storage of the fish.

Porsby et al. 2008 determined that the numbers of L. monocytogenes decreased, but were not eliminated in a liquid smoked salmon product when vacuum packed and stored at 5oC over a period of 20 days. However, although the study comprehensively assessed a number of stages of the cold smoking process, the fillets were all artificially inoculated.

The effectiveness of vacuum packaging (VP) was compared against modified atmosphere packaging (MAP) and assessed using general microbial indicators, physico-chemical changes in the product and sensorial changes in sliced cold smoked rainbow trout during storage of 60 days (Iacumin et al. 2017). MAP packaged fish showed lower microbial numbers compared with VP packaged samples throughout the entire storage period. The average count of generic mesophiles in the VP packaged product exceeded 6 log cfu/g after 60 days of storage. Total volatile base-nitrogen (TVB-N) concentrations increased with time in both packaging types and exceeded that set limit of 40 mg N/100 g after 45 days. The authors concluded that the manufacturer-set shelf life of 60 days was too long for both packaging types. Sensory evaluation favoured MAP over VP, which was the same ranking as the microbiological determinations. Consequently, a final conclusion of the study was that MAP (70% N2, 30% CO2) was more effective than VP for the preservation of cold smoked trout against spoilage (N. B. spoilage microbiota not L. monocytogenes). A review of comparisons between VP and MAP (Lovdal, 2015) concluded there was a wealth of information that supported MAP with high [N2] or [CO2] as being the more reliable for extending shelf life compared with VP.

In summary, it is apparent that naturally present L. monocytogenes can multiply under vacuum packing conditions on smoked fish. Although heavy salting in combination with some smoke residues can significantly delay growth and possibly even cause partial L. monocytogenes death, vacuum packing in itself is not an effective way of controlling L. monocytogenes growth or eliminating it entirely during cold storage prior to consumption.

Carbon dioxide modified atmosphere packing

In the EU, modified atmosphere packaging (MAP) for foods are considered to be food additives and were originally approved under Annex 1 of 95/2/EC and updated by EC 1333/2008. MAP is permitted for use subject to the conditions and labelling requirements stipulated in directives EC 95/2 and 1333/2008. In addition, Regulation 33 of the Food Labelling Regulations 1996 sets out that: “A food, the durability of which has been extended by means of its being packaged in any packaging gas authorised pursuant to Council Directive 89/107/EEC, concerning food additives for use in foodstuffs intended for human consumption, shall be marked or labelled with the indication packaged in a protective atmosphere.”

Nilsson et al (1997) report that carbon dioxide modified atmosphere (70% CO2 [E290] and 30% N2 [E941]) packed cold-smoked salmon artificially inoculated with a seven-strain cocktail of L. monocytogenes had an eight-day lag before L. monocytogenes growth. Control samples packed under vacuum showed a five-log increase in vacuum packaging under the same conditions. However, the CO2-mediated prevention of growth was a temporary effect because after 27- and 44-days post inoculation, there were three log and four log increases in L. monocytogenes numbers respectively.

Increasing the CO2 concentration to 100%, in combination with either of the bacteriocins nisin or ALTA 2341, was shown to completely suppress growth of L. monocytogenes artificially inoculated onto salmon at both refrigeration and abuse temperatures (Szabo and Cahill 1999).

Based solely on subjective sensory determinations, Muratore and Licciardello (2006) reported that cold smoked sliced swordfish had a much shorter half-life of 12 days when packed in a modified atmosphere (5% O2, 45% CO2, 50% N2) compared with 42 days when vacuum packed. The study showed no correlation between the shorter sensory shelf-life of the modified atmosphere packed smoked fish and increases to the total aerobic mesophilic counts. Bacterial numbers for both treatments did not significantly increase until at least 15 days storage which was 3 days past the sensory shelf life (Muratore and Licciardello, 2006).

Changes in L. monocytogenes numbers during refrigerated storage of artificially contaminated rainbow trout fillets packaged in air (control), vacuum and various modified atmospheres (MAP) have been studied (Yilmaz et al. ,2009). The MAPs used were 50% CO2 and 50% N2 (MAP A); 80% O2 and 20% CO2 (MAP B) and 2.5% O2 in combination with 7.5% N2 and 90% CO2 (MAP C). Over a storage interval of 18 days, the L. monocytogenes populations multiplied from an initial concentration of 104 cfu/g to 107 cfu/g in air and to 106 cfu/g in vacuum packaging. Yilmax and colleagues observed that modified atmosphere packaging did not eliminate L. monocytogenes from rainbow trout fillets. However, all three MAPs retarded the growth of L. monocytogenes at 4oC to some extent. There was little difference between the antibacterial effects of each MAP, with each having the lowest count at different points along the experimental time course. After 18 days MAP B had the lowest count (5.5 log cfu/g fish), but MAP C was only slightly higher (6.7 Log cfu/g fish). The air control contained 7.5 Log cfu/g fish. Whether the differences in the treatments were significant is not stated and it is therefore difficult to draw firm conclusions from the interesting, but ineffectively reported, study (Yilmaz et al., 2009).

Although treatment of smoked fish with a CO2-based MAP can likely delay L. monocytogenes exponential growth, there is a significant problem with CO2 packing which also applies to a number of other MAP gases. MAP requires an increased volume of packaging to maintain, for example, an effective CO2 to product ratio. The extra volume of packaging produces a downstream reduction in sustainability through the supply chain. Some analysts have excluded this intervention as not feasible based on the increase in packaging volume (Hansen et al. 2009). However, there is potential to circumvent this problem by utilizing a CO2 emitter to maintain an adequate level of CO2 inside a low volume package during storage (Hansen et al. 2009). To date no studies have investigated the use of CO2 emitters as an intervention for L. monocytogenes and smoked fish. However, using the CO2 emitter approach, Hansen and colleagues found that the total bacterial numbers on raw fish remained below 50 cfu/g over the first 14 days of storage at 0.1oC. After two weeks however, the numbers of total aerobes had begun to increase slowly. The average numbers of total aerobic mesophiles had increased by around two log units by 28 days. By comparison, controls without the CO2 emitters and an air atmosphere had bacterial numbers of around seven logs.

Modified atmosphere packs using elevated concentrations of CO2 have issues to be resolved before they could be used as an intervention for the control of L. monocytogenes on smoked fish. It is likely that CO2 emitters would need to be used to prevent the logistical cost issues associated with increased packaging volumes. In addition, CO2 does not eradicate L. monocytogenes or prevent its growth on smoked fish. Under conditions of perfect refrigeration, high concentration CO2 MAP at best delays the proliferation of L. monocytogenes by up to seven days. In order to be used as a reliable intervention, CO2 MAP would need to be used in combination with another treatment such as a bacteriocin (e. g. nisin).

Product packed in MAP must be appropriately labelled.

Sodium nitrite primarily added to cold smoked fish as an intervention has been shown to inhibit the growth of Clostridium botulinum (Pelroy et al. 1994). Over a period of 40 days at 5oC L. monocytogenes growth could be inhibited by a combination of 3% or 5 % NaCl along with 190-200 ppm sodium nitrite when packaged in gas-permeable film or under vacuum. However, when the experiments were performed at 10oC, growth was detected after as little as 5 days of the 40-day study. A combination of nitrite and NaCl was effective at low inoculums (10 cfu/g) of L. monocytogenes, but when greater numbers (327 cfu/g) were used, the inhibition was not so pronounced. Even with a low inoculum, growth was detected when fillets were incubated at 10oC with L. monocytogenes numbers of 105-106 cfu/g detected after 30 days (Pelroy et al. 1994). These results reinforce the importance of having uncontaminated starting material followed by adequate refrigeration for effective L. monocytogenes control.

Lerfall and Østerlie (2013) investigated the use of sodium nitrite as a preservation agent for cold smoked salmon. The work was general and undertook chemical evaluations of the consequences of nitrite treatment in terms of carcinogenic N-nitrosamine formation, as well as organoleptic assessment of colour changes and texture. It was concluded that the use of nitrite in cold-smoke processing of farmed Atlantic salmon changed the product colour by increasing the isomerisation of astaxanthin, a lipid-soluble antioxidant that confers the pink colour to salmon flesh. Lerfall and Østerlie (2013) also reported improved microbiological stability and increased shelf life of the product. An important finding of the study was that there was no evidence for significantly increased concentrations of nitrosamines or peroxides after treatment. Nitrosamines are carcinogens formed when nitrite is applied to organic materials and their consumption in products such as cured bacon is a risk factor for some cancers (Michaud et al. 2006). In June 2017, EFSA published two scientific opinions on nitrites and nitrates added to food.

In summary, nitrite used without other inhibitory agents, or in combination with high salt concentrations, has limited use as an intervention L. monocytogenes proliferation on smoked fish. Nitrite is effective at postponing, but not eliminating, the commencement of exponential growth by L. monocytogenes only when the initial starting numbers of the bacteria are very low (~10 cfu/g fish). Although nitrite cannot be used in the EU at the current time for fish, it is widely used to cure other meats in Europe. A single study (Lerfall and Østerlie, 2013) found no evidence for significantly increased concentrations of nitrosamines or peroxides in nitrite-treated fish. Consumption of these compounds are risk factors for some cancers.

In the USA, sorbates can be used as food additives since they are generally recognised as safe (GRAS). In the UK there are regulations in force transposed from the EU. Sorbates are subject to limits on their final concentrations, as preservatives for some foods including semi-preserved fish and salted fish products (The Food Additives, Flavourings, Enzymes and Extraction Solvents (Amendment etc.) (EU Exit) Regulations 2019).

The effect of pre-treatment with potassium sorbate on L. monocytogenes populations during the storage of freshly caught red mullet and carp in Greece by Tassou et al. (2004). Gutted fish were dipped for 60 seconds in sterilized water containing L. monocytogenes at a concentration of 3 x 105 cfu/ml, then placed the fish in a solution of 5% (w/v) potassium sorbate with or without hot water at 60oC (Tassou et al., 2004). The fish were subsequently packed in an aerobic or in a modified atmosphere (40% CO2, 30% O2, and 30% N2) and stored at 0-1oC. In the control (untreated) fish, the numbers of L. monocytogenes increased from 4.8 to 6.5 log cfu/g. By comparison, L. monocytogenes numbers on the potassium sorbate-treated fish stored aerobically remained at 4.0 log cfu/g over a 15-day period. Similar results were seen with red mullet and when using sorbate in combination with hot water (Tassou et al. ,2004). However, when the fish were stored in a modified atmosphere, the difference between controls and potassium sorbate treated samples were not so marked. The authors speculate that an inhibitory effect to L. monocytogenes growth was conferred by the MAP although previous studies have shown little practical benefit for MAP as a control measure for L. monocytogenes. No indications were given on what effect, if any, there was on the sensory quality of the fish.

There are a number of issues relating to the studies of Tassou et al. (2004) which make potassium sorbate treatment unsuitable for use as an intervention for L. monocytogenes. Firstly, it is likely that both the use of MAP and the sorbate caused the observed bacteriostatic effect on raw fish; but the inhibitory contribution of each treatment was not clear. Secondly, the hot water treatment of 60oC used by the researchers was likely to have denatured at least some of the protein in the fish flesh potentially causing organoleptic changes to the finally smoked product. For these reasons, potassium sorbate treatment was not assessed as a suitable intervention for L. monocytogenes-contaminated raw fish.

Potassium sorbate in a combination with other antimicrobials has also been found to be inhibitory to L. monocytogenes (Ye et al, 2008). Ye et al (2008) tested a range of GRAS preservatives which included nisin, sodium lactate, sodium diacetate, potassium sorbate and sodium benzoate. Effectiveness testing was for the chemicals individually and also in combination. The study concluded that when used in conjunction with chitosan films, a combination of sodium lactate and potassium sorbate was the most effective of the antimicrobials assessed. The combination prevented the growth of L. monocytogenes for over eight weeks at 4˚C (Ye et al. 2008).

In a study undertaken in the United States by Neetoo and colleagues (2008a), potassium sorbate (PS) used on its own was shown not to significantly inhibit growth of a 12-strain cocktail of L. monocytogenesin broth culture at its (EU and US) legal concentration limit of 0.3% (w/v; Neetoo et al. 2008a). However, when PS was used in combination with nisin (0.00125% and 0.0025% w/v) and sodium diacetate (0.125 and 0.25%w/v) no growth in broth was detected. Although results from broth-based studies can sometimes bear little relation to real-world observations in commercial processing facilities, Neettoo et al. (2008) were also able to show that a combination of 0.00125% Nisin/0.15% PS could suppress the growth of L. monocytogenes in smoked salmon fillets over three weeks storage at 4oC.

In summary, sorbate is not able to effectively eliminate or prevent the growth of L. monocytogenes on smoked fish during storage of the product. Sorbate has been shown to be effective when used in combination with lactate or nisin. However, both nisin (Nilsson et al., 1997) and lactate (Pelroy et al., 1994) on their own showed significant efficacy at preventing the proliferation of L. monocytogenes. Thus, the observed inhibitory effects for the combination sorbate treatments are probably more due to the lactate and nisin than the sorbate. Consequently, sorbate is not assessed as being a particularly useful intervention for L. monocytogenes control on smoked fish.

Electrolysed oxidising (EO) water is generated by passing an electrical current through a weak solution of sodium chloride dissolved in tap water. The electrolysis generates hydrogen gas and hydroxide radicals at the cathode. At the anode, the chlorine ions from the salt are neutralised and form chlorine gas. If the chlorine at the anode is reacted with hydroxide at the cathode, hypochlorite (the active agent in bleach) is formed. If the pH of the solution is lowered, hypochlorous acid is formed (Fabrizio et al. 2002). EO water is used to describe solutions of hypochlorite, hypochlorous acid and mixtures of these two antibacterial agents.

Shiroodi et al. (2016) evaluated EO water for the control of L. monocytogenes on cold smoked salmon. The Shiroodi study evaluated the antimicrobial activity EO water as a pre-treatment method before the commencement of the processing of cold-smoked salmon. Sensory and textural quality determinations of the final product were undertaken. Raw Atlantic salmon fillets were inoculated with L. monocytogenes at a concentration of 6 x 105 cfu/g and treated with EO water at 20oC, 30oC or 40oC and at three different exposure durations of 2, 6 or 10 min before cold smoking the fish. EO water applied at 40oC did not cause significant change to the organoleptic properties of the fish. However, the same treatment also lowered the L. monocytogenes counts by almost 3 log cfu/g.

Ovissipour et al. (2018) also compared the effect of different EO solutions, including acidic electrolysed water (AEW) and neutral electrolyzed water (NEW), alone and in combination with mild thermal processing (50oC, 55oC, 60oC, 65oC) at different exposure times (2, 6, 10 min) on the reduction of L. monocytogenes on Atlantic salmon fillets. The study was undertaken by the same research group that reported the Shiroodi et al. (2016) results. Changes in the proteins of the L. monocytogenes and also the salmon muscle were assessed using Fourier transform infra-red (FTIR) spectroscopy. The highest observed bacterial reduction was 5.6 log10 cfu/g observed at the highest thermal processing temperature (65oC) after the longest time (10 min) on salmon treated with NEW. In general, NEW was a better AM than AEW, particularly when combined with mild thermal processing. NEW caused less alteration of the protein structures in the fish. EO water is also known as hypochlorous acid. The authors of this study were based in the USA, where EO water is GRAS and legal for use on seafood at concentrations up to 60 ppm (Dewi et al., 2017).

Heir et al. (2019) evaluated the fate of a cocktail of L. monocytogenes inoculated onto cold smoked salmon that was prepared using acetic acid during the dry salting stage of processing and stored under vacuum for one month. The end product was both organoleptically and microbiologically assessed. Compared to a control not treated with vinegar, the acetic acid treatment resulted in increased lag times and reduced growth of the L. monocytogenes. The degree of L. monocytogenes inhibition was dependent on the type of contamination (i. e. between slices or a non-sliced salmon application), the concentration of acetic acid used, and the storage duration and temperature (4 or 8oC) as well as the intensity of smoking. For sliced salmon, growth was inhibited (but L. monocytogenes was not killed) at 4oC storage for 29 days when 1% acetic acid was used. Under the same conditions, there was a three-log increase in L. monocytogenes numbers in the untreated control fish. Under conditions of temperature abuse (8oC) there was a two-log increase in L. monocytogenes when 1% acetate was used, compared with an increase of 5-6 logs for the control. A taste panel had no consistent preference for salmon prepared with and without acetic acid. The paper’s conclusion was acetic acid had robust listeriostatic effects which could be used in combination with listericidal UV-C light treatment to further reduce the listeria-risks of this ready-to-eat food product category.

Baptista et al 2010 reviewed historical publications dealing with organic acids such as citric acid, acetic acid, and lactic acid generally. In the USA, these acids are all considered to be GRAS. The review concluded that a mixture of organic acids was most effective at controlling bacterial spoilage because different acids dissociated under different conditions, and acid dissociation is involved in the antimicrobial mechanism of action. Despite the conclusions of the Heir (2019) study, the review also discussed that a major barrier to the use of organic acids in cold smoked fish was the associated taint and the denaturation of fish muscle protein by low pH. Furthermore, there were no reports of effective L. monocytogenes control by organic acids in the review of historical information. Organic acids are used extensively and legally across a number of food sectors including bovine carcass washing in the EU. In general, acids reduce populations and injure bacteria, but do not entirely destroy them. Sub-lethally injured cells can recover from acid treatments.

Bacteriocins are antimicrobial proteins secreted by some bacterial species which tend to inhibit only closely related bacterial species (Cotter et al. 2005). Bacteriocins are secreted by bacteria primarily as part of their niche protection strategy. Nisin is a polycyclic peptide bacteriocin from Lactococcus lactis subsp. lactis widely used as a food preservative (Delves-Broughton et al., 1996). The use of nisin as a food preservative has been widespread in the UK since the 1960s (Delves-Broughton et al., 1996). Nisin (E234) is authorised for food preservation in the European Union by 95/2/EC as updated by EC 1333/2008. Nisin is the only bacteriocin currently permitted for use in the EU and its use is restricted to only ripened and processed cheeses, dairy-based puddings, clotted cream and mascarpone. Specifications for nisin are laid down in Directive EC 1129/2011. Permitted concentrations for nisin are product-specific and the highest concentration allowed is 12.5 mg/kg for cheese. Nisin may be present in food as a consequence of the direct addition of purified bacteriocin, or it can be secreted from naturally present or inoculated lactic acid bacteria.

Several bacteriocins have been identified as efficiently reducing L. monocytogenes numbers and are discussed here with the caveat that different L. monocytogenes strains vary in their susceptibility to nisin (Rasch and Knochel, 1998). Approximately 50 units/g nisin resulted in the survival and growth of L. monocytogeneson smoked salmon (Nilsson et al., 1997). Low concentrations of nisin remained ineffective even when used in combination with other preservatives such as salting and CO2 packing. However, raising the nisin concentration to 500 or 1000 U/g of cold-smoked salmon inoculated with L. monocytogenes delayed, but did not prevent growth in vacuum-packs stored at 5oC. At higher concentrations of nisin to CO2 packaged cold-smoked salmon resulted in an initial one to two log reduction of L. monocytogenes numbers followed by a lag phase of 8 and 20 days in salmon with 500 and 1000 U nisin/g, respectively (Nilsson et al. 1997). An elongated lag was also reported by Szabo and Cahill (1999) who also investigated growth of L. monocytogenes in nisin-treated smoked salmon packed under vacuum.

The effects of Nisin, sodium lactate or their combination (1:1) injected into rainbow trout at an industrial scale before the smoking process as well as into the finished smoked product has been reported (Nykänen et al. (2000). Both nisin and sodium lactate were observed to inhibit the growth of L. monocytogenes in refrigerated smoked fish. However, in combination the two compounds acted synergistically (Nykanen et al. 2000) and when injected into finished product, decreased the numbers of L. monocytogenes from 3.26 to 1.8 log cfu/g over 16 days of storage at 8oC. The numbers of L. monocytogenes remained almost constant (4.66-4.92 log cfu/g) for 29 days at 3oC in those samples injected before smoking with nisin and sodium lactate. However, any sensory implications of the intervention were not recorded and so it is currently unknown if taint or texture changes are a consequence of the treatment.

Neetoo et al. (2008b) examined how L. monocytogenes growth changed when nisin-coated plastic films were used to vacuum-pack cold smoked salmon. In control (non-coated) samples, the numbers of L. monocytogenes artificially inoculated onto pressed disks of salmon pâté or cold smoked salmon fillets grew from 500 cfu/cm2 to approximately 1 x 107 cfu/cm2 during storage for 58 days at 4oC. Pâté samples wrapped in plastic coated with 500 IU of nisin/cm2 displayed a decreased rate of L. monocytogenes growth but eventually reached the same level as the control. When 2000 IU of nisin/cm2 of film was used, the growth of L. monocytogenes on the pâté was inhibited over the 58 days of 4oC storage. In addition, when the increased concentration of nisin was used, growth on pâté was inhibited for more than 35 days at a higher storage temperature of 10oC. Inhibition was also found when the cold smoked salmon covered with a plastic film coated with 2000 IU of nisin/cm2 was inoculated with L. monocytogenes at a concentration of 3 logs cfu/cm2 and stored at 4oC for 43 days. However, in contrast to the pâté, no inhibition was noted when the salmon inoculated with the higher level of L. monocytogenes was stored at a higher temperature (10oC) which is routinely used to mimic temperature abuse during storage (Neetoo et al., 2008b).

In Chile, Concha-Meyer et al. (2011) reported the effects of wrapping smoked salmon in an alginate film containing nisin, followed by refrigerated storage under vacuum. It is difficult to directly compare the Concha-Meyer et al. (2011) and Neetoo et al. (2008b) studies because Concha-Meyer provides the concentration of nisin used in the liquid prior to film formation, whereas Neetoo et al. (2008b) reports the nisin concentration per cm2 of film. It is likely the Concha-Meyer work (100 IU nisin/ml) used less nisin than the Neetoo et al. (2008b) study. Correspondingly, Concha-Meyer et al. (2011) reported an initial inhibition of the growth of an L. monocytogenes population of 103 cfu/g for two weeks. Between 14 and 28 days, however, L. monocytogenes numbers increased exponentially to more than 105 cfu/g.

Lebow et al. (2017) report their findings from another investigation of the effects of nisin and low-temperature HPP on L. innocua (used as a non-pathogenic surrogate for L. monocytogenes). Spoilage, organism growth, colour, sensory properties of the treated fish and ‘peelability’ were objectively assessed. Cold smoked salmon fillets with and without nisin (10μg/g) were inoculated with a 3-strain L. innocua cocktail before vacuum-packaging and freezing at -30oC. High-pressure was applied to the fish in an ice slurry within an insulated sleeve. Initial experiments indicated that nisin and HPP for 120 s at 450 MPa (N450) and 600 MPa (N600) were most effective against L. innocua, and so those conditions were used for subsequent assessments. L. innocua in N450 and N600-treated CSS was reduced 2.63 and 3.99 log cfu/g, respectively. L. innocua and spoilage growth were not observed in HPP-treated CSS during 36 d storage at 4oC. The low temperature application of HPP was beneficial in reducing the lightening of the muscle colour caused by the treatment. Sensory evaluation indicated a preference for HPP salmon by over 60% of the panelists (P<0.05). Peelability of sliced CSS was reduced by HPP (P<0.05). Nisin in combination with low-temperature HPP was effective in controlling L. innocua in CSS and improved consumer organoleptic experiences of the product.

Chen et al. (2020) investigated whether L. monocytogenes being stressed in a fish processing environment might upregulate stress responses and change any susceptibility to nisin. In vitro experiments were undertaken that were appropriate mimics for in vivo conditions. It was concluded that pre-exposure to a slightly acid environment, exposure to high salt concentrations and a sublethal exposure to quaternary ammonium sanitiser activated stress control responses that were likely to provide cross-protection against a subsequent nisin treatment of L. monocytogenes on cold-smoked salmon. An important finding of the work was that challenge studies that use lab cultured strains and nutrient-rich media may overestimate the efficacy of nisin as a control strategy for L. monocytogenes on cold-smoked salmon. Cultured cells are not stressed by a lack of nutrients and not expressing genes that confer protection against nisin.

Bacteriocins have been isolated from a range of Lactobacilli (Ghalfi et al. 2006). A bacteriocin secreted by Lactobacillus curvatus CWBI-B28 was shown to be inhibitory to L. monocytogenes growth in broth and on cold smoked salmon fillets (Ghalfi et al. 2006). A range of different application methods including direct addition of the bacterial strains to fillets, spraying fish with partially purified bacteriocin and packaging in a bacteriocin-coated film have been explored. Packing films were optimally coated by heat inactivation of a culture containing Lactobacillus producer cells and manipulation of cell solution to an acidic pH. The coated film treatment showed promise as an effective control measure. Numbers of L. monocytogeneson film-packaged cold smoked salmon declined from 2 log cfu/cm2 fish to a level of less than the detection limit of 5 cells per cm2 after three days at 4˚C (Ghalfi et al. 2006). Furthermore, the film prevented any subsequent increase in the L. monocytogenes numbers throughout 22 days of storage at 4oC.

A bacteriocin-producing Carnobacterium divergens strain is known to secrete an intervention potential. The C. divergens strain used for the study secreted the M35 bacteriocin (Tahiri et al. (2009). Both the culture and the purified M35 bacteriocin were assessed as inactivators of L. monocytogenesin cold smoked salmon. A 2.6 log cfu/g reduction in the numbers of L. monocytogenes was observed for up to 10 days of storage in samples treated with the C. divergens culture. Purified divergicin M35 (50 mg/g), or crude culture supernatant showed reductions of a single log cfu/g at the beginning of storage. However, the anti-listerial activity of the supernatants lasted for 15 days compared to only three days for purified bacteriocin. Colour and texture were not significantly affected by any of the treatments.

Banos et al. (2016) assessed the ability of the LAB enterocin AS-48 to control L. monocytogenes in fish during storage at 4oC. In addition, the effectiveness of a dual treatment of enterocin and Listex P100 (PhageGuard) bacteriophage was evaluated. Listex has been assessed as GRAS (generally recognised as safe) by the USDA and FDA. The safety and efficacy are not quite recognised by EFSA. Currently Listex P100/PhageGuard is not permitted for fish in the EU or UK. A single treatment of bacteriocin reduced L. monocytogenes inoculated onto raw hake or raw salmon by around three logs after seven days of refrigerated storage. The Listex P100/PhageGuard treatment also reduced inoculated numbers of L. monocytogenes significantly, but less effectively, than the bacteriocin, in both raw fish types. A dual treatment of bacteriocin and phage eliminated listeria from hake and salmon fillets in less than two days. In cold smoked salmon, a combined treatment of AS-48/Listex P100/PhageGuard reduced listeria to below detection levels from 1 to 15 d. However, after 15 days, listeria that were injured and not killed by the dual treatment recovered, and a small number of cells could be detected by enrichment.

When evaluating any antimicrobial, the likelihood of the development of resistance by the target organism must be considered. Studies determining high levels of resistance to carnobacteriocin B2 for fish isolates of L. monocytogenes (Nilsson et al. 2006) suggest that if lactic acid bacteria are used as an intervention for L. monocytogenes associated with smoked fish; it may be a better strategy to use lactic acid strains which do not secrete bacteriocins to prevent the development of resistance. However, it has also been argued that cocktails of different bacteriocin-producing strains would reduce the likelihood of the emergence of resistant strains (Galvez et al. 2010).

Also covered by the Baptista et al. (2020) review is LAB and the bacteriocins they export into their environments. The primary mechanism of action of many bacteriocins (pore formation in bacterial plasma membranes) is discussed as is the development of resistance to bacteriocins. A common route to resistance is the secretion of protease which degrades the short chain proteins that form bacteriocins. In addition, the potential difference (typically -70mV in bacteria) can be altered by using alternative metabolic pathways and terminal electron acceptors. Such changes in the voltage across the bacterial plasma membrane make it more difficult for the bacteriocins to insert into the membrane as a precursor to pore formation. The most positive papers summarised by the Baptista review describe only an inhibition of growth or in a small number of studies, small declines in L. monocytogenes populations.

In summary, bacteriocins such as nisin show promise as interventions for L. monocytogenes associated with smoked fish. The principal drawbacks for bacteriocin use are that large quantities of the antimicrobial (500-1000 U/cm2 fish) are required for effective L. monocytogenes control. Furthermore, it has already been shown that L. monocytogenes can develop a resistance to individual bacteriocins. Thus, in order to ensure prolonged effectiveness either a cocktail of strains or bacteriocins or an alternative strategy such as bacteriocin in combination with lactate would be required for the prevention of resistance in the longer term. It is likely there are legal barriers to the use of bacteriocins for the control L. monocytogenes populations on fish.

Lee et al. (2012) describes the laboratory-scale manufacture of antimicrobial (AM) films from mustard meal. Mustard meal is a by-product of a bio-fuel process. Two main treatments for the films were assessed. The first was inoculation of smoked salmon fillets with L. monocytogenes followed by wrapping in the film (WIF), the second was wrapping the fish in a film and then inoculating (WTI). For the WIF treatment, the mustard film inhibited more than 4.0 log cfu/g of L. monocytogenes at the lower storage temperatures assessed for 7 days. There was a temperature-influenced growth of L. monocytogenes after 7 days for the WIF treatment. At 5oC inhibition was maintained for 28 days, at 10oC there was no significant growth for 14 days, and at 15oC significant growth had occurred by 7 days. Inhibition was more pronounced when the coating was applied before inoculation because the L. monocytogenes had to physically penetrate the wrapping. Under this treatment, at 15oC, L. monocytogenes was isolated from the fish by day 14. At 10oC and 5oC, isolations from fish were by day 21. Although the films show some activity when used to wrap fish fillets inoculated with laboratory-cultured L. monocytogenes, these films are a long way from being suitable for commercial use. Over the last ten years, the same research group have reported a number of similar films manufactured from other processing by-products such as whey powder from the dairy industry. To date, none have been developed to usable products that have been adopted by the food manufacturing industries. Although promising, it is considered unlikely this study would form the basis of workable intervention at the current time. The paper has been included into this review, largely to keep the technical content current.

Erkan and Yesiltas (2016) evaluated another AM film for any ability to impede the growth of a range of generic, non-pathogenic indicators on hot smoked rainbow trout fillets. The effect of fish coated with an alginate-based coating containing 1% thymol (thyme essential oil) (w/w) was compared with untreated controls during refrigerated (2oC) storage over seven weeks. The study concluded that the film was effective in preserving hot smoked trout fillets from lipid oxidation and generic microbial growth. The observed reductions were to a range of indicators and the film inhibited, rather than curtailed bacterial growth. No apparent consideration was made of taint, although thymol is a strong flavouring.

Albertos et al. (2017) investigated ground up or extracted olive leaves incorporated into edible films for the inhibition of L. monocytogenes in cold-smoked fish. The use of such films for the decontamination of food has questionable legality in the EU. Olive leaf powder (OLP) and its water/ethanol extract (OLE) were found to inhibit L. monocytogenes growth in a previous study by the authors that used only agar diffusion tests. The current paper showed antimicrobial activity of the films increased with increasing OLE concentration in their formulations. A film formulation with around 5% OLE, significantly reduced the growth of L. monocytogenes (NB: but did not completely inhibit L. monocytogenes) inoculated onto fish during 6 days’ storage. In conclusion, the film most likely could not be used legally in the EU at present. The film slowed the growth of lab cultured L. monocytogenes inoculated onto fish, but L. monocytogenes growth still occurred over 6 days storage.

Moreno et al. (2017) created edible biofilms from a blend of corn starch (CS) and bovine gelatine (BG) with and without ethyl lauroyl arginate (LAE) as antimicrobial agent. The films were assessed for their ability to curtail general bacterial growth and L. innocua (as a surrogate for L. monocytogenes). The films with and without LAE showed antilisterial activity in in vitro tests. Packaging of marinated salmon samples in these films greatly reduced the total viable counts, which remained below the legal limit after 45 storage days at 5oC. Nevertheless, films were not effective at controlling weight (i. e. water) loss of salmon samples during the extended cold storage.

The senior author of a publication by Baek and Song (2018) has a history of the (experimental) development of antimicrobial films. A brief search for patents and previously described film products did not reveal any obvious further development after publication for previously reported films. The current paper focused on film development using processed seaweed (Gracilaria vermiculophylla) as a raw material. The seaweed was rich in agar and agarose and a gel created from them was used as the foundation for the film; glycerol and sorbitol were added as plasticisers and zinc oxide nanoparticles (ZnONP) were added to create an antibacterial film. The films were assessed for antibacterial activity and their ability to block UV light. The UV consideration was made because blocking UV light slows lipid oxidation of fish fat and the development of off-flavours caused by rancidity. The films were applied as a packaging material for smoked salmon. UV-visible light transmission was decreased by the addition of ZnONPs. Exposure to the film for 9 h caused a decrease to the numbers of L. monocytogenes (and also Salmonella Typhimurium). For films containing 5% zinc nanoparticles, a five-log reduction was observed in the numbers of inoculated L. monocytogenes. The transmission of UV light was also blocked by the film. The authors believed their results indicated that the films containing zinc oxide nanoparticles had potential as an active food packaging material to maintain the quality of smoked salmon. No work was undertaken to determine if the zinc oxide leached into the product. It is also unlikely films of this type would be permitted in the EU.

The aim of a study by Benabbou et al. (2018) was to evaluate the effectiveness of a film containing chitosan and a bacteriocin called divergicin 35 on L. monocytogenes on cold smoked salmon. The salmon was inoculated with L. monocytogenes and treated with the chitosan film and/or bacteriocin before storage at 4-8oC for three weeks. In some replicates, the film reduced L. monocytogenes to below the detection limit (<50 cfu/g) and on average kept total counts below 104 cfu per g compared to 109 cfu per g in untreated control samples. In contrast, the effectiveness of the bacteriocin applied in solution was poor. The film showed suppressive activity for at least three weeks, but the bacteriocin in solution was not significantly different from an untreated control at 14 days storage. The film additionally, positively protected desirable organoleptic traits for the fish such as flesh colour and firmness. Although the bacteriocin was not particularly effective, the authors concluded divergicin-loaded chitosan film may have benefit for the bio-preservation of cold-smoked fish.

Preservatives such as nitrates, sulphates and sorbate may not be approved in all countries likely to import cold smoked fish and are specifically forbidden in some EU member states where they are banned from being used to enhance the shelf-life of smoked fish (Matamoros et al. 2009). Therefore, there are drivers that have caused some authors to investigate novel methods for reducing L. monocytogenes contamination of fish as alternatives to traditionally used food preservatives. Lactic acid bacteria (LAB) are promising candidates because some members of the group secrete multi-factorial antimicrobial compounds such as lactic acid, hydrogen peroxide and bacteriocins.

Although LAB are part of the natural biota associated with meat, fish, milk and cheese, if they are cultured and added to foods then there may be regulations which apply. A summary of uses and whether regulations apply are shown as Table 3. The EC regulations which could potentially prevent or restrict the use of LAB on fish are Regulation EC 2015/2283 on novel foods and novel food ingredients, Directive 1333/2008 on food additives, Directive EC 1334/2008 on flavourings for use in foods and Directive 2002/46/EC on food supplements, (Wessels et al. 2004).

Table 3: Potential LAB functions in food and regulatory categories (compiled from Wessels et al. 2004)

Although LAB have a widely acknowledged ability of inhibiting the growth and multiplication of a variety of food spoilage organisms (Wessels et al., 2004), Nilson et al. (1999) assessed the ability of LAB in inhibiting L. monocytogenes growth. The Nilson study used non-identical strains of Carnobacterium piscicola that had been found to dominate the biota of refrigerated vacuum packed stored cold smoked salmon along with a strain of Lactobacillus sake. Two strains of Carnobacterium piscicola were inoculated at a concentration of ~2 × 106 cfu/g onto salmon slices which were also inoculated with L. monocytogenes (~2 × 102 cfu/g) before storage at 5oC. On salmon slices without the lactic acid bacteria, L. monocytogenes grew to 3 × 108 cfu/g (Nilsson et al. 1999). When co-inoculated, both strains of Carnobacterium piscicola were able to inhibit the growth of L. monocytogeneson refrigerated salmon slices for over 40 days (Nilsson et al. 1999). The very large inoculum numbers used by the study meant that the anti-listerial effect may have been due to competition for nutrients. In order to determine if that was the case, external nutrients were applied to the salmon slices. No significant differences between the results for the additional nutrient samples and the standard nutrient samples were observed strongly suggesting that the basis of the inhibition was not nutritional. The sensory profile for Carnobacterium piscicola inoculated salmon was reported to be the same as for untreated cold smoked salmon samples. However, the strain of Lactobacillus sake produced undesirable changes in flavour (Nilsson et al. 1999).

The Nilsson study (1999) established that C. piscicola was antagonistic to L. monocytogenes at high numbers. However, later studies have also reported that C. piscicola is a promising antagonist of L. monocytogenes on smoked salmon even when inoculated at lower numbers. Yamazaki et al. (2003) found a strain of C. piscicola (~104 cfu/g) was able to effectively control the growth of L. monocytogenes (~103 cfu/g) when co-inoculated on to salmon aerobically at 4oC and 12oC. The lactic acid-producing strain was able to completely inhibit the growth of L. monocytogenes for over 20 days. At 20oC, the growth of L. monocytogenes was still significantly reduced compared with the control (not inoculated with C. piscicola) samples.

Vescovo et al. (2006) found that a combination of two LAB strains (Lactobacillus casei and Lactobacillus plantarum) inoculated at 6 logs cfu/g onto cold smoked salmon achieved a reduction in the counts of L. innocua of 3.2 logs compared to the control during storage of the product under vacuum for 30 days at refrigerated temperatures.

As discussed by the pioneering studies of Nilsson et al. (1999), the antagonistic nature of competing bacteria may be multi-factorial and therefore not as easily broken as can be the case with the addition of a single controlling substance. When 57 strains of L. monocytogenes were checked for their resistance to three strains of antagonistic Carnobacteria, none were found to be resistant (Brillet et al., 2004). Some grouping of more and less sensitive L. monocytogenes strains could be made, but generally, the antagonistic abilities of the Carnobacteria were maintained in situ on vacuum-packed, cold-smoked salmon refrigerated for 4 weeks (Brillet et al., 2004). The most effective strain of the three Carnobacteria was C. divergens V41, a strain previously identified by Duffes et al. (1999) as being able to antagonise L. monocytogenes when added in co-culture on sterile homogenized cold smoked salmon. Further, later studies were performed to determine the sensory characteristics of cold smoked salmon inoculated with the V41 strain. A trained panel could notice a slight taste difference, but it was felt to be too small for untrained consumers to notice (Brillet et al. 2005).

Tome et al. (2006) found that only 41% of LAB strains exhibited an antibacterial effect on L. innocua in a plate assay. However, the authors believed that their cultures were very competitive and that they may provide additional protection against the growth of L. monocytogenes. An interesting observation of the Tome study was that for cold smoked salmon, storage at refrigeration temperatures (5oC) in vacuum packaging allowed the LAB strains to outcompete other bacteria thereby potentially putting L. monocytogenes under competitive as well as selective pressure. A later study (Tome et al. 2007) identified the most effective conditions (6h dry salting with sugar, 6 h of drying and 2 h of smoking) for growth of lactic acid bacteria in vacuum packaged cold smoked salmon. The research concluded that the growth of lactic acid bacteria that are anti-listerial can be enhanced by the appropriate selection of processing parameters.

Concha-Meyer et al (2011) reported the effects of wrapping smoked salmon in an alginate film containing two different LAB strains and nisin followed by refrigerated storage under vacuum. For the in vivo section of the paper, only the two LAB strains in combination or in combination with nisin were evaluated. After demonstrating in vitro activity for the films by plate assay, there was a further evaluation using salmon inoculated with laboratory-cultured L. monocytogenes to 104 cfu/g. The salmon were examined on a weekly basis to determine numbers of L. monocytogenes. After 28 days, when the study was terminated, control salmon pieces covered with alginate film without inhibitors showed an increase of 2.4 logs for L. monocytogenes. Films with either of the LAB strains or a combination of both strains and nisin had a bacteriostatic effect, with the numbers of L. monocytogenes unchanged over the storage. Twenty-eight days, is toward the maximum typical shelf life for cold smoked salmon. Although the study used a single lab-cultured strain of L. monocytogenes, the results demonstrate that the impregnated films can effectively inhibit L. monocytogenes growth on salmon during refrigerated storage.

In Spain, Montiel et al. (2013) have reported extended lag and minor inhibition of L. monocytogenes as a consequence of brining salmon with sugar and salt in combination with commercial preparations of bacteriocin-secreting Pediococcus acidilactici, mixed cultures of Lactobacillus spp. or Lactobacillus curvatus under conditions of temperature abuse. Initial inoculation of fresh salmon was to roughly four log cfu/g fish, with storage at 8oC for seven days. Pediococcus acidilactici in combination with salt and sugar was able to prevent the multiplication of L. monocytogenes over the duration of the study. Lactobacillus curvatus in combination with salt and sugar showed a small decline of 0.4 log cfu/g in L. monocytogenes numbers over the course of the week. In contrast, the mixed LAB culture in combination with salt and sugar showed increases in L. monocytogenes numbers after three days and an increase of more than one log cfu/g after one week, which was similar to the salt and sugar control treatment. An untreated control increased by around three log cfu/g over the seven-day duration of the experiment.

Aymerich et al. (2019) also investigated the effect of LAB on L. monocytogenes contaminating fish. Different smoked fish types were used representing a range of lipid concentrations, water activities, salt and phenolics. Three LAB strains were investigated, all of which LABs produced organic acid and bacteriocins. Two of the LABs were isolated from smoked fish, with the third originally isolated from red meat. The fish isolate LABs reduced the growth of L. monocytogenes but did not inhibit it. However, the meat-borne starter culture, L. sakei CTC494 inhibited growth and the authors speculate, improved the food safety of cold-smoked salmon. Aymerich et al. (2019) are based in Spain and are aware of the EU restrictions on decontamination treatments. The issue is briefly discussed and Aymerich approach to the issue is novel and disregards the antimicrobial treatment. A quote from the authors is “… considering current EU legislation, L. sakei CTC494 was the only bioprotective culture that enabled the product to be changed from category 1.2 (RTE food able to support the growth of L. monocytogenes) to category 1.3 (RTE food not able to support the growth of L. monocytogenes) (European Commission, 2005), thus categorising it at a lower risk.” In essence, the authors propose ignoring the EU legal restrictions on decontamination treatments and focussing on the use of LAB changing the food categorisation from supporting the growth of L. monocytogenes to not supporting the growth of L. monocytogenes. The described approach is the opinion of the study authors only and has never been tested in court as a defence against the enforcement of the regulations used in the UK.

Wiernasz et al. (2020) also undertook research designed to extend the shelf life of gravadlax using LAB. The main consideration of the study was prevention of organoleptic deterioration and spoilage, although some experiments were designed to assess the effect of LAB on food safety by considering L. monocytogenes. Sliced gravadlax, with a commercial shelf-life of 21 days, was purchased from a French processor and inoculated by spraying with lactic acid bacteria at a concentration of 106 cfu/g. The impact of the inoculation on the gravadlax microbiota, sensory properties, biochemical properties and volatilome (volatile flavour compounds) was followed for 25 days of storage at 8oC in VP. The antimicrobial activity of the LAB was also assessed against L. monocytogenes. Six LAB strains were assessed. Three of the strains were competitive in the fish and became the dominant bacteria on the fillet by inhibiting the growth of the spoilage microbiota and also suppressed L. monocytogenes. These three strains conferred their own flavour and volatilome sensory characteristics to the fish. The remaining three LAB did not significantly impact on the fish flavour, the spoilage microbiota or significantly suppress L. monocytogenes. However, one of the largely innocuous LAB was able to preserve the sensory quality past 25 days. The study authors concluded that C. maltaromaticum SF1944 and V. fluvialis CD264 both had potential as bioprotective cultures to ensure salmon gravadlax microbial safety and sensorial quality, respectively.

In summary, some LAB are able to prevent the growth of susceptible strains of L. monocytogenes for around three weeks when refrigeration is effective. Some LAB strains can prevent L. monocytogenes multiplication for at least seven days under refrigeration abuse conditions. As a general rule, increasing the number of LAB strains used tends to extend the amount of time that L. monocytogenes growth is restricted. Although flavour alterations in the product will be influenced by the LAB strains in use, there are some strains which do not cause significant changes to texture or flavour. Subject to compliance with the regulations specified at the start of this section, the use of multiple strains of LAB rather than purified bacteriocin, would be expected to help reduce the likelihood of resistance in susceptible populations of L. monocytogenes.

Trisodium phosphate (TSP; E339) has been approved for use as a food additive within the EU for a narrow range of foods. Currently, TSP cannot be used for the control of L. monocytogenes in fish.

In the United States, TSP is generally regarded as safe for raw food (Mu et al. 1997). In the US, TSP is routinely used for the decontamination of raw poultry meat by a patented process which involves the immersion of post-chill whole poultry in a 10% (w/v) solution for 15 minutes (Mu et al. 1997). However, TSP is a less effective decontaminant for fish. After storage for 9 days at 4oC, pond-reared rainbow trout immersed in either a 10% or a 20% solution of TSP did not show significant reductions in total psychrotrophic counts or numbers of artificially inoculated L. monocytogenes when compared with control samples which had been immersed only in tap water (Mu et al. 1997).

Sodium lactate (E325; SL) is approved for food use in the European Union for a narrow range of foods. Lactate cannot currently be used for the control of L. monocytogenes in fish.

The growth of L. monocytogenes in cold smoked fish in comminute (minced) raw salmon was mixed with a range of concentrations and combinations of sodium lactate, sodium chloride, and sodium nitrite. Samples of fish were inoculated with 150 L. monocytogenes cells, vacuum-packaged in oxygen-impermeable film and stored at either 5oC or 10oC. Periodically, the samples were tested to determine the numbers of L. monocytogenes until the end of the product’s shelf life (50 days). Pelroy and colleagues (1994) determined that sodium lactate exhibited a concentration-dependent ability to prevent the growth of listeria, but that it did not inactivate L. monocytogenes.

Furthermore, the inhibition of growth was enhanced by the presence of nitrite and/or increased concentrations of NaCl. The prevention of L. monocytogenes growth was more pronounced at 5oC where total inhibition of L. monocytogenes growth was achieved for up to 50 days in the presence of 2% sodium lactate and 3% (water-phase) NaCl. At 10oC, total inhibition was achieved for up to 35 days by 3% sodium lactate and 3% (water-phase) NaCl, or by 2% sodium lactate in combination with 125 ppm sodium nitrite and 3% water-phase NaCl.

The antimicrobial effects of different concentrations of potassium lactate (in combination with sodium diacetate) were also evaluated by Yoon et al. (2004) for control of L. monocytogenes. The use of potassium lactate (PL) plus sodium diacetate mixture at all tested dilutions completely inhibited the growth of L. monocytogenes on smoked salmon stored at 4oC during 32 days of storage (Yoon et al. 2004).

Researchers (Vogel et al. 2006) found that a combination of potassium lactate (2.1%) and sodium diacetate (0.12%) delayed the growth of L. monocytogenes for up to 42 days in vacuum packed cold smoked salmon stored at 10oC. This procedure did not affect the quality of the product and the authors suggest that it is a suitable technology to prevent the growth of L. monocytogenes.

Mejlholm and Dalgaard (2007) reported that MAP gravad cold-smoked salmon with the addition of 0.15% (wt/wt) sodium diacetate (SD) prevented the growth of L. monocytogenes for more than 40 days at 8oC, whereas the addition of 0.15% (wt/wt) diacetate reduced the growth rate of the pathogen in MAP cold-smoked Greenland halibut. This difference between the two types of products was explained by a higher content of naturally occurring lactate in cold-smoked salmon (0.77 to 0.98%, wt/wt) than in cold-smoked Greenland halibut (0.10 to 0.15%, wt/wt).

Ye et al. (2008) assessed the effectiveness of SL incorporated into a chitosan-coated plastic film. During an initial evaluation, chitosan-coated plastic film containing sodium lactate at 4.5 mg/cm2 was assessed as effective at inhibiting the growth of a cocktail of 5 x 105 cfu L. monocytogenes /cm2 fish at 20oC for ten days. When the lactate-infused film was tested at refrigeration temperature, it completely inhibited the growth of L. monocytogenes on smoked salmon for at least 6 weeks. The authors concluded that chitosan-coated plastic films containing 4.5 mg/cm2 lactate can potentially assist the smoked-salmon processing industry in their efforts to control L. monocytogenes.

Follow on work by Ye et al. (2011) assessed the effectiveness of SL and SD at various concentrations in combination and also evaluated a commercial formulation of SL and SD called Opti.Form. The Ye et al. (2011) study incorporated the SL and SD into edible films made from alginate, k-carrageenan, pectin, gelatin and starch. The films were used to wrap cold smoked salmon fillets that had been inoculated with a five-strain cocktail of laboratory-cultured strains to a concentration of 500 cfu/cm2 fish surface. In the first section of the study, the wrapped, inoculated fish slices were frozen to -18oC for 6 days, thawed and then stored at 22oC for 6 days. 2.4% SL and 0. 25% SD in gelatin was the most effective inhibitor of L. monocytogenes. However, there was benefit for 2.4% SL and 0.125% SD in alginate and 2.4% SL and 0.25% SD in starch. In the control samples L. monocytogenes grew from 2.9 log cfu/cm2 to 7.5 log cfu/cm2. For the most effective gelatin treatment, L. monocytogenes numbers reduced slightly to 2.7 log cfu/cm2. The alginate and starch treatments were both lower than four log cfu/cm2. Thus, there was effective inhibition for some of the films for fish exposed to the temperatures used by the study.

In the second section of the study, cold-smoked salmon slices were inoculated with L. monocytogenes to 3200 cfu/cm2 fish surface, coated with alginate, gelatin or starch film with or without incorporating the antimicrobials and stored frozen at -18oC for 12 months (Ye et al. 2011). Every two months, samples were removed from the freezer and stored at 4oC for 30 days before testing. In unwrapped controls, the numbers of L. monocytogenes were not significantly affected by frozen storage at -18oC for 12 months. Immediately after thaw, the control film samples without SL and SD showed a reduction of around one log cfu/cm2. Addition of antimicrobials to the films further decreased the population of L. monocytogenes by around two log10 cfu cm-2 after 12-month frozen storage. There was no statistical difference among any of the antimicrobial treatments at each sampling point of the frozen storage. In the absence of prior frozen storage of the fish, L. monocytogenes in the control samples grew steadily from to more than six log cfu/cm2. For the unfrozen samples, all three edible coatings containing antimicrobials inhibited the growth of L. monocytogenes and slightly reduced its population during the 30-day storage at 4oC. When inoculated control samples stored frozen for 2–8 months were compared with samples stored frozen for 10–12 months there were differences in final numbers of L. monocytogenes after the refrigerated storage. Consequently, the study authors concluded that >2 months frozen storage and antimicrobial edible coatings represent an effective intervention to inhibit the growth of L. monocytogenes on cold-smoked salmon.

Freezing is a complex process for microorganisms, with the rate of freezing being a key consideration for cellular death (Harrison et al 2013). Once frozen, there is generally little change to bacterial populations during an extended freeze as was reported in the first section of the study by Ye et al (2011). Although the Ye et al. (2011) study reported results at odds with the seminal studies of Guyer and Jemmi (1991) neither studies report freeze rates, which makes direct comparison difficult. Furthermore, different L. monocytogenes strains, culture conditions and test methodologies were used for each study, which exacerbates the comparison issues.

Kin et al (2012) assessed the effects of potassium lactate (PL) and potassium acetate (PA) on L. monocytogenes -contaminated catfish fillets in combination with liquid and wood smoke. In the EU, PA and PL are currently not permitted for use as an antimicrobial on fish and so the inclusion of the Kin et al. (2012) study is to maintain an accurate technical overview of potential fish decontamination treatments, rather than a list of useable interventions. Kin prepared catfish fillets by tumbling in either a treatment of PA (0.25%) and PL (0.58% w/v), or a phosphate buffer control. Post tumble treatments applied were no smoke, gaseous wood smoke or one of two smoke condensates. After the fish had been processed, they were coated with a three-strain cocktail of lab cultured L. monocytogenes at a concentration of 104 cfu/g fish. The fate of the L. monocytogenes was followed during refrigerated storage at 4oC. After 14 days, the most effective treatments for inhibition of L. monocytogenes were PA, PL and wood smoke, followed by PA, PL and liquid smoke. None of the treatments applied completely inhibited L. monocytogenes. There was around 0.5 log cfu/g growth in the PA, PL and wood smoke treated fish and over 1 log cfu/g in the PA, PL and liquid smoke treatment. In the controls, L. monocytogenes growth exceeded 3 log cfu/g. There was no evidence of any significant organoleptic changes as a consequence of the PL and PA treatments. Organoleptic assessment of the acceptability revealed a preference for PA, PL and wood smoke treated fish. Whilst the Kin et al (2012) study shows there is some benefit in treating with PA and PL in combination with smoking, the intervention did not completely control the L. monocytogenes hazard and so the study findings are likely to be of significant benefit only is used in combination with other intervention(s).

Tang et al (2013) investigated the effect of SD and PL, nisin and binary combinations on L. monocytogenes populations. PL and SD in combination was synergistically anti-listerial, and the most effective treatment because the combination extended the lag before exponential growth to around 8-10 days on average for both L. monocytogenes in inoculated broth and inoculated onto smoked salmon. In addition, the PL and SD combination reduced the maximum growth rate, and final cellular densities on salmon.

Valo et al. (2020) examined a process more usually associated with meat processing to see if it was applicable for fish processing. The process stage was an application of purified smoke condensate (PSC) to fish using an atomiser mist. The use of PSC for fish is legal within the European Union. However, the paper also studied the effect of PSC in combination with potassium lactate, which may not be allowed in the EU. PCS applied as a mist reduced aerobic plate counts and numbers of lactic acid bacteria. The authors claim consequently that their process had a potential to enhance the shelf life of CSS. The study found that the phenolics in the PSC and the lactate had only a minor influence on growth inhibition. What was more important for hindering the growth of the spoilage bacteria was water activity and pH. The authors concluded that drying and optimising the PCS application were key to obtaining a high-quality CSS, with a long shelf life.

Edible alginate coatings containing lactate and diacetate have been appraised as possible suppressors of L. monocytogenes growth on cold smoked salmon fillets (Neetoo et al., 2010). The study incorporated a range of concentrations of SL and SD into five edible (approved in the EU food-grade) coatings (alginate, κ-carrageenan, pectin, gelatin or starch). A range of concentrations of sodium lactate were used, either alone or in combination with SD. The researchers applied the coatings onto the surface of cold smoked salmon slices inoculated with L. monocytogenes at 500 cfu/cm2 before storage at room temperature (~22oC) for six days. Despite the high ambient temperatures, the alginate coatings were able to contain L. monocytogenesgrowth at approximately 500 cfu/cm2 and were the most effective carriers for delivering L. monocytogenes growth-inhibiting antimicrobial compounds. Neetoo et al., (2010) concluded that alginate-based coatings containing lactate and diacetate could make a meaningful food safety impact for fish consumed without cooking. In combination, lactate and diacetate delivered in an alginate coating are bacteriostatic, checking the growth of L. monocytogenes and thereby enhancing the microbiological safety of filleted and sliced smoked salmon.

In summary, lactate in combination with chitosan, diacetate, high salt and/or nitrite appears able to delay or slow the growth of L. monocytogenes for extended periods of at least 30-40 days. If high salt concentration is used, the temperature needs to be meticulously maintained at 5oC or lower to constrain L. monocytogenes growth. Two studies have shown that acetate in combination with diacetate or lactate in combination with acetate does not appear to alter the flavour or texture of smoked fish and thus these treatments show real potential as a workable intervention measure. Lactate has been approved for use in the EU as a bactericidal (decontamination) wash on only some foods of animal origin. At the present time, approval has not been extended to fish.

Acidified sodium chlorite (ASC) is an antimicrobial treatment and it has been evaluated as a possible control intervention for the growth of L. monocytogenes artificially inoculated onto raw whole salmon and salmon fillets (Su and Morrissey 2003). Article 3(2) of Regulation (EC) No 853/2004 of the European Parliament and of the Council lays down specific hygiene rules for food of animal origin and provides a legal basis to permit the use of a substance other than water to remove surface contamination from products of animal origin. However, approval of the use of any substance in compliance with 853/2004 requires an evaluation of the decontamination chemical to be undertaken by an Expert Panel of the EU and for the chemical to be assessed as safe. We are unable to find any evidence that a specific evaluation for ASC has been undertaken by an EU Expert Panel for use with fish. Thus, it is likely that ASC cannot currently be used for the control of L. monocytogenes in fish.

However, a number of antimicrobial washes including ASC were evaluated by an EFSA expert panel to determine whether the use of such washes as chicken carcass decontamination treatments would result in the emergence of resistant bacterial strains (EFSA, 2008). The panel found no evidence that resistance would occur. However, the opinion for chicken is only a small amount of the considerations required for a complete safety assessment. It is important to note that the FAO believe if is possible under some circumstances that the breakdown of ASC can result in chlorate. Regulation (EU) 2020/749 imposes maximum residue levels for chlorate in or on certain products.

Su and Morrissey (2003) spray washed fish in 50ppm ASC solution for one minute before storage on ice (fillets) for one week or frozen storage (whole fish) for one month. Changes in L. monocytogenes numbers associated with the salmon were followed every two days for the fillets and at the end of storage for the whole salmon (Su and Morrissey, 2003). The ASC treatment did not reduce L. monocytogenes numbers on the salmon skin for the frozen whole fish. On the fillets, the initial ASC wash reduced populations of L. monocytogenes by roughly 0.5 logs. However, when the fillets were stored in frozen water ice, the L. monocytogenes numbers increased slowly. If the fillets were stored in frozen ASC ice, L. monocytogenes growth still occurred although it was slower compared with the standard water ice (Su and Morrissey, 2003).

In summary, ASC treatment of raw fish had a small benefit in reducing the growth of L. monocytogenes, but it only slowed and did not completely stop L. monocytogenes multiplication. Although Su and Morrissey report that there was no difference in colour of ASC treated salmon, they also say that further studies would be required on sensory characteristics such as taste, before ASC could be adopted as a modest control measure.

Bacteriophages are specific viruses which infect and damage bacterial cells. Eukaryotic cells such as human cells are unaffected by bacteriophages. The use and mode of action of bacteriophages in food production is contentious and was the subject of a series of reviews and expert opinions by EFSA. In combination, these reviews and opinions summarised the complex regulations and consumer hazards which may apply to foodborne bacteriophages. The most recent opinion and evidence summary has been published (EFSA, 2016). The EC held a consultation on phages in summer 2016 but did not act either to ban or approve bacteriophage products and treatments. In essence, the current situation is that the use of bacteriophage as decontaminant on products of animal origin is currently not permitted in the EU and neither has the use of bacteriophage as food additives been authorised.

The use of bacteriophages for the control of L. monocytogenes on fish have been evaluated (Soni and Nannapaneni 2010) and it was determined that the commercially-available bacteriophage LISTEX P100 (EBI Food Safety, Wageningen, Netherlands) could inhibit the growth of L. monocytogenes on raw salmon fillet tissue over a 10 day storage at 4oC. The phage was able to lyse L. monocytogenes numbers to as low as 0.3 log cfu/g smoked fish, while levels in untreated control samples were as high as 2.6 log cfu/g. In addition to providing a two-log reduction to L. monocytogenes numbers, Soni and Nannapaneni (2010) believed that, in contrast to most anti-Listerial treatments, application of the phage preparation in a saline wash helped to prevent any deteriorations in the quality or aesthetic appearance of the smoked product.

Phage-based control of L. monocytogenes has been shown to cause beneficial reductions to L. monocytogenes numbers in some foods (Soni et al., 2012; Holck and Berg, 2009; Guenther et al., 2009). However, the surface of cold smoked salmon has been shown to contain widely diverse strains of L. monocytogenes (Guenther et al., 2009). In studies designed to determine the importance of strain variation, the ability of bacteriophage to control two smoked fish isolates of L. monocytogenes was assessed. In brief, phage was added to the surface of cold-smoked salmon previously inoculated with a mixture of L. monocytogenes strains each at a concentration of 103 cells per fillet. The food was refrigerated for six days at 6oC before examination. The results were that one strain had been reduced by over two log units but that the other L. monocytogenes population was not significantly affected by the phage treatment. In other commodities (not cold smoked salmon) both phages were shown to be effective, this difference is likely due to the differing physiochemical properties of the food product itself. The most difficult foods to treat with phage are those with an uneven and large surface area (fish, meat, and seafood), which physically limits the distribution of phage particles in order to reach all bacterial targets.

Vongkamjan et al. (2013) undertook analyses of historical isolates from a smoked fish plant in order to determine ability to persist in a processing plant and susceptibility to bacteriophage. The 141 L. monocytogenes strains used for the study were isolated between 1998 and 2009. Over that 11-year period, strains with identical ribotypes and pulsotypes were observed. In essence the result means either near continuous re-colonisation of the plant for 11 years, or more likely persistence for the same time. Eleven years persistence is longer than any previous report. In addition, Vongkamjan et al. (2013) determined L. monocytogenes strain susceptibility to lytic bacteriophages. Challenge studies were undertaken using high numbers of L. monocytogenes (>105 cfu). The authors observed up to four logs reduction could be achieved when using cocktails of mixed phage. There was a wide range of likelihoods from 5% to 95% that any individual strain would be susceptible to phage. A single L. monocytogenes strain was unaffected by any of the phages evaluated. One of the important messages from the Vongkamjan et al (2013) study was that susceptible populations of L. monocytogenes grew back after exposure to bacteriophage. However, the regrown populations were typically less susceptible when re-exposed to the same phage. Although more work is required to confirm the results, it seems possible that any extensive use of L. monocytogenes bacteriophage would select for resistant populations and become increasingly ineffective over time.

van Nassau et al. (2017) investigated the use of PlyP40, Ply511, or PlyP825 phage endolysin (which degrade the peptidoglycan in bacterial cell walls during the lytic phase of phage propagation) in combination with high hydrostatic pressure processing as a control intervention for L. monocytogenes. It seems likely that the van Nassau et al. (2017) study was an initial appraisal using broth, to test if meaningful reductions were possible using the combination approach, before moving onto food. There was a killing synergy observed for L. monocytogenes when HPP and endolysin was used, compared with the effect of just a single treatment. More than a five-log kill was possible with the combination treatment of endolysin and HPP. The study concluded that an application of endolysins did not only substantially increased the bactericidal effect of high pressure, but it also enabled the inactivation of bacterial cells at much lower pressure levels.

The aim of a study by Misiou et al. (2018) also assessed the activity of phage endolysins, in combination with HPP. The paper is from the same research group as the van Nassau et al. (2017) publication. Both publications used the same endolysins and the same pressures of HPP using the same equipment. The van Nassau work was undertaken in broth, the Misiou study was undertaken using selected foods including cold smoked salmon. The test organisms were a cultured cocktail of L. monocytogenes. Determinations of lethality and fate after storage were performed. For smoked salmon, a minimum pressure level of 500 MPa (without lysins) was required to inactivate Listeria cells in smoked salmon and at that pressure, the authors described the impact as ‘still rather limited’ and in the order of 1.5 log cycles. L. monocytogenes in the other (non-aquatic) foods that were assessed were more susceptible. When HPP was applied in combination with lysins, the smoked fish was still resistant to the treatments. In contrast, the other assessed foods were more effectively decontaminated. By the authors’ own admission, the process described by Misiou et al. (2018) was not effective enough to be useful in a commercial setting for the decontamination or an extension of shelf life for cold smoked fish.

It is likely that bacteriophages and enzymes derived from them cannot currently be legally used for the control of L. monocytogenes in fish in the EU. More information and a larger comprehensive review of bacteriophages as decontaminants in seafood has been published by Lasagabaster et al. (2020).

A short study by Pennisi et al. (2020) evaluated ultrasound at different temperatures as decontamination treatment for fish. Four strains of L. monocytogenes were lab cultured and applied to fish as a cocktail at a high concentration of 108 cfu/g. Ultrasound of the fillets was applied using an ultrasonic water bath set to 20oC, 25oC, 30oC, 40oC or 50oC, with sonication for 5, 10 or 15 minutes. Thermosonication treatments between 40oC and 50oC for 5, 10 or 15 minutes proved to be the most effective decontamination treatments, which were able to remove roughly 2 log cfu/g from the fish. The ultrasound treatments did not significantly alter the organoleptic properties of the fish.

Lysozyme treatment is a recognised preservative under regulation EC 1169/2011. Lysozyme can trigger allergic reactions in sensitive individuals. Thus, the inclusion of even food grade lysozyme, sourced from chicken egg albumen, is required to be specifically labelled. Currently, lysozyme cannot be used for the control of L. monocytogenes in fish.

Datta et al. (2008) studied the potential of lysozyme treatment in combination with alginate (polysaccharides extracted from brown seaweed) to prevent the growth of L. monocytogenes inoculated onto smoked salmon. The effects of exposure to nisin, two forms of lysozyme isolated from oysters and egg white were assessed both individually and in combination with calcium alginate applied to the surface of the fish. None of the treatments caused a statistically significant reduction to the growth of L. monocytogenes over a 35-day period at 4oC compared with the untreated controls. The result of using nisin alone was in broad agreement with the other studies dealt with earlier in this review (Nilsson et al. 1997; Szabo and Cahill 1999; Nykanen et al. 2000; Neetoo et al., 2008b) in that there was a weak and not significant inhibition of growth. A slight reduction in the numbers of L. monocytogenes was observed by combining nisin or either of the two lysozymes with calcium alginate, however, the best reduction was achieved by adding both nisin and lysozyme to calcium alginate where there was over a two-log reduction compared with the untreated controls. The study of Datta et al. (2008) also found broadly comparable reductions in numbers of Salmonella anatum using the combination of calcium alginate, nisin and lysozyme.

Essential oils are complex mixtures of hydrophobic compounds extracted from fruits and vegetables. Essential oils are produced mostly as flavourings for food manufacture, however some essential oils, such as allicin extracted from fresh garlic are also potent antimicrobials (Ankiri and Mirelman, 1999).

In the EU, essential oils in foods are subject to Regulation 1334/2008/EC as a consequence of being flavouring agents. Regulation 1334/2008/EC restricts the use of biologically active agents in flavourings absolutely on the basis of food type and also by capping the maximum permissible concentration of each active compound in those foods which are approved. Thus, essential oils may be legal as ingredients and permitted as additives.

Lin et al. (2004) have investigated a combination of oregano and cranberry essential oil extracts, both of which contain antimicrobial phenolic compounds. In summary, the essential oils were very good at reducing numbers of L. monocytogenes. Numbers of L. monocytogenes were measured after the addition of both oregano and cranberry extracts alone and in combination. No decreases in L. monocytogenes populations were observed on raw cod slices kept at 4oC and individually treated with either cranberry or oregano extract. However, when a mixture (75% oregano:25% cranberry; v/v) was assessed, there was a significant three log unit decrease observed over an 8-day refrigeration period (Lin et al. 2004). The decrease was observed earlier if lactic acid was used to acidify the extracts to pH 6.0. Under the acidified conditions, a comparable three log decrease was observed at as early as four days under refrigeration. No information was provided on what sensory effects the addition of such extracts might have or the potential increases to manufacturing costs.

Yuan et al. (2017) screened 125 plant and herb extracts looking for compounds that were able to inhibit L. monocytogenes. One of the plants screened was an extract from Cinnamomum javanicum leaves (cinnamon). The phenolic content of the extract was 78mg phenolics/g extract and the MIC of the unrefined extract was 0.13 mg/ml. The effectiveness of the AM was dependent on the strains of L. monocytogenes that were exposed. The more resistant L. monocytogenes strains required such a high concentration of the extract, that the salmon flesh was discoloured. GC-MS revealed that around one quarter of the unrefined extract consisted of eucalyptol. Both crude extract and eucalyptol induced significant membrane damages in exposed L. monocytogenes. In summary, the extract’s impact on L. monocytogenes was unpredictable and strain dependent. No consideration was made that plant extracts have variable concentrations of AM between different extract batches. Although the conclusion from the authors was the extract holds promise as a control treatment for L. monocytogenes, the reality is far more work would be required before that might be considered to be the case. The principal active agent eucalyptol would be required to purified (or synthesised) and its concentration standardised before more realistic effectiveness assessments could be made.

Baptista et al. (2020) is a review of preservation treatments for fish. The review covers essential oils (EOs) and plant extracts (PE) purified from herbs such as thyme, cumin and cinnamon and typically concentrated by distillation. The review identifies broad classes of antimicrobial compounds in EOs and summarises their mechanisms of action. There was also a section describing how to optimise the concentrations of EOs in plants during cultivation. The review included an exceptionally comprehensive table of around 100 papers that have shown activity of EOs/PEs in relation to seafood spoilage. The table also covered packed PEs, in for example, artificial antimicrobial films. A wide range of fish and other seafoods were covered. In brief, a summary of the EO activity table is that some PEs can have some impact in reducing or inhibiting the growth of some bacterial populations (including L. monocytogenes). There were no papers however that reported the effective destruction of any spoilage bacteria. Although PEs and EOs are extracted from edible plant material, if the primary reason for their inclusion in food is decontamination, they are probably still subject to EU regulations.

Other AM treatments covered by the Baptista et al. (2020) review include chitosan, a short chain sugar polymer that is antibacterial and extracted from exoskeletons of shellfish such as lobster. Chitosan is commonly applied to seafood in the form a low pH solution. Typically, the acid used to achieve the low pH is acetic acid (vinegar) and so it would not be suitable for cold smoked salmon because it would introduce a taint to the product.

β-phenylethylamine (PEA) was evaluated as an organic antimicrobial for the control of L. monocytogenes and as a new general food safety intervention (Muchaamba et al. 2020). The study focussed on food generally, although smoked fish was one of the model foods used to assess in vivo activity. The authors assembled a collection of test L. monocytogenes strains that were sourced from clinical infections and food-related samples. The minimum inhibitory concentration (MIC) of PEA against L. monocytogenes was initially measured in broth and agar growth media to determine if there was any AM activity. After encouraging initial results, Bologna sausage and cold smoked salmon were used as model systems to validate the in vitro growth media findings. PEA had a growth inhibitory and bactericidal effect against L. monocytogenes both in in vitro experiments as well as on the sausage and smoked salmon. The MICs ranged from 8 to 12.5 mg/ml; typical bacterial growth inhibitors are applied in the mg/ml range. There was evidence that growth of that type that formed biofilms was impeded by PEA. The authors concluded that the AM PEA might be an additional hurdle to help curtail L. monocytogenes growth thereby increasing food safety. No assessment was made of the organoleptic consequences of using PEA. No discussion of the legal issues preventing the use of PEA in the EU was undertaken. The authors were based in Switzerland, which is not an EU member state. Irsfeld et al. (2013) have reviewed PEA, which is naturally present in and can be synthesised by human metabolism. Using PEA in food may not be advisable and should probably be subject to robust scrutiny by regulatory authorities. PEA is a human neurotransmitter and “low or high concentrations of PEA may be associated with specific psychological disorders” … [including] “attention deficit hyperactivity disorder, depression, and schizophrenia” (Irsfeld et al. 2013).