perforation | traçabilité | EMAP | bactériologie    

U. S. Food and Drug Administration
for Food Safety and Applied Nutrition
September 30, 2001

Analysis and Evaluation of Preventive Control Measures for the Control and
Reduction/Elimination of Microbial Hazards on Fresh and Fresh-Cut Produce

Chapter VI

Microbiological Safety of Controlled and Modified Atmosphere Packaging of Fresh and Fresh-Cut Produce


This chapter addresses the use of modified atmosphere packaging and controlled atmosphere packaging for the preservation of fresh produce. There have been great technological advances in this area of preservation, particularly as it refers to improving the quality and shelf-stability of highly perishable food products, such as produce. However, when using these technologies, careful attention must be paid to the effect on the survival and growth of pathogenic organisms. This chapter focuses on food safety aspects of packaging technologies that are either commercially available or under investigation.

4. Conclusions

  • Oxygen, CO2, and N2, are most often used in MAP/CAS. Among them, CO2 is the only one with a direct antimicrobial effect, resulting in an increased lag phase and generation time during the logarithmic phase of growth. Although other gases such as nitrous and nitric oxides, sulphur dioxide, ethylene, chlorine, as well as ozone and propylene oxide have been investigated, they have not been applied commercially due to safety, regulatory, and cost considerations.
  • The recommended percentage of O2 in a modified atmosphere for fruits and vegetables for both safety and quality falls between 1 and 5%, although the oxygen level will realistically reach levels below 1% in MAP produce.
  • The concern when using MAP for fruit and vegetables arises from the potential for foodborne pathogens, which may be resistant to moderate to high levels of CO2 (< 50%), to outgrow spoilage microorganisms, which may be susceptible to the modified atmosphere.
  • It is generally believed that with the use of permeable films, spoilage will occur before toxin production is an issue; MAP of produce, however, should always incorporate packaging materials that will not lead to an anoxic package environment when the product is stored at the intended temperature.
  • The background microflora is largely responsible for the spoilage of fresh produce and can vary greatly for each product and storage conditions. The elimination or significant inhibition of spoilage organisms should not be practiced, as their interaction with pathogens may play an integral role in product safety.
  • Edible films for use in MAP systems is an active area of research. However, as with other MAP, they can create a very low O2 environment where anaerobic pathogens such as C. botulinum may thrive. Antimicrobial compounds that can be incorporated into the coating are also being currently investigated.
  • There have been many studies investigating the migration of antimicrobials such as sodium benzoate, benzoic acid, propionic acid, and potassium sorbate from coatings into food. It appears that the most advantageous use of these films for antimicrobial properties would be the formation of a monolayer lipid and sorbic acid film, or a bilayer film composed of a hydrophilic base layer coated with a thin layer of lipid containing sorbic acid. The main issue involves the production of coatings with good surface tension that will stick to produce.
  • Successful control of both product respiration and ethylene production and perception by MAP can result in a fruit or vegetable product of high organoleptic quality; however, control of these processes is dependent on temperature control. Along the whole food continuum, that is, processing, storage, transportation and retailing, one needs to maintain optimum temperatures. Maintaining proper storage temperatures is often most difficult at retail level.
  • Currently, there is concern with psychrotrophic foodborne pathogens such as L. monocytogenes, Y. entercolitica and A. hydrophila, as well as non-proteolytic C. botulinum, although clearly a number of other microorganisms, especially Salmonella spp., E. coli O157:H7 and Shigella spp., can be potential health risks when present on MAP produce.
  • The success and microbiological safety of MAP is dependent on controlled low temperature storage and the product's characteristics.
  • Only two MAP produce products, coleslaw mix and ready-to- eat salad vegetables, have been implicated in foodborne illness outbreaks of botulism and Salmonella Newport , respectively.


1. Introduction

Over the past 20 years, there has been an enormous increase in the demand for fresh fruit and vegetable products that has required the industry to develop new and improved methods for maintaining food quality and extending shelf life (see Chapter I). Due to the complexities involved with produce, that is, varying respiration rates which are product and temperature dependent, different optimal storage temperatures for each commodity, water absorption, byproducts, and so on, many considerations are involved in choosing an acceptable packaging technology. One of the areas of research that has shown promise, and had success, is that of modified atmosphere packaging (MAP). This technique involves either actively or passively controlling or modifying the atmosphere surrounding the product within a package made of various types and/or combinations of films. In North America , one of the first applications of this technology for fresh-cut produce was introduced by McDonald's (Brody 1995), which used MAP of lettuce in bulk-sized packages to distribute the product to retail outlets.

The major factors responsible for extending the shelf life of fruits and vegetables include: careful harvesting so as not to injure the product, harvesting at optimal horticultural maturity for intended use, and good sanitation (Moleyar and Narasimham 1994; Lee and others 1996). When these are practiced, the implementation of optimum storage conditions through modified atmospheres can be quite effective at maximizing the shelf life and quality of the product.

A modified atmosphere can be defined as one that is created by altering the normal composition of air (78% nitrogen, 21% oxygen, 0.03% carbon dioxide and traces of noble gases) to provide an optimum atmosphere for increasing the storage length and quality of food/produce (Moleyar and Narasimham 1994; Phillips 1996). This can be achieved by using controlled atmosphere storage (CAS) and/or active or passive modified atmosphere packaging (MAP). Under controlled atmospheric conditions, the atmosphere is modified from that of the ambient atmosphere, and these conditions are maintained throughout storage. Examples of this type of storage and the commercial systems available are listed in Table VI-1. MAP uses the same principles as CAS; however, it is used on smaller quantities of produce and the atmosphere is only initially modified. Active modification occurs by the displacement of gases in the package, which are then replaced by a desired mixture of gases, while passive modification occurs when the product is packaged using a selected film type, and a desired atmosphere develops naturally as a consequence of the products' respiration and the diffusion of gases through the film (Moleyar and Narasimham 1994; Zagory 1995; Lee and others 1996). The numerous film types used in MAP are listed in Table VI-2, and some commercially available MAP systems are listed in Table VI-3.

Oxygen, CO2, and N2, are most often used in MAP/CAS (Parry 1993; Phillips 1996). Other gases such as nitrous and nitric oxides, sulphur dioxide, ethylene, chlorine (Phillips 1996), as well as ozone and propylene oxide (Parry 1993) have been suggested and investigated experimentally. However, due to safety, regulatory and cost considerations, they have not been applied commercially. These gases are combined in three ways for use in modified atmospheres: inert blanketing using N2, semi-reactive blanketing using CO2/N2 or O2/CO2/N2 or fully reactive blanketing using CO2 or CO2/O2 (Parry 1993; Moleyar and Narasimham 1994).

Normally, the concentration of O2 in a pack is kept very low (1-5%) to reduce the respiration rate of fruits and vegetables (Lee and others 1995). Reducing the rate of respiration by limiting O2 prolongs the shelf life of fruits and vegetables by delaying the oxidative breakdown of the complex substrates which make up the product. Also, O2 concentrations below 8% reduce the production of ethylene, a key component of the ripening and maturation process. However, at extremely low O2 levels (that is, <1%), anaerobic respiration can occur, resulting in tissue destruction and the production of substances that contribute to off-flavors and off-odors (Lee and others 1995; Zagory 1995), as well as the potential for growth of foodborne pathogens such as Clostridium botulinum (Austin and others 1998). Therefore, the recommended percentage of O2 in a modified atmosphere for fruits and vegetables for both safety and quality falls between 1 and 5% (Table VI-4). However, it is recognized that the oxygen level will realistically reach levels below 1% in MAP produce. It is generally believed that with the use of permeable films, spoilage will occur before toxin production is an issue; MAP of produce, however, should always incorporate packaging materials that will not lead to an anoxic package environment when the product is stored at the intended temperature. This recommendation should be qualified, however, by saying that all films are permeable to oxygen to some degree; the difference pertains to the rate of gas transfer through the film, with some films allowing greater transfer rates than others. Moreover, the elimination or significant inhibition of spoilage organisms should not be practiced, as their interaction with pathogens may play an integral role in product safety. A number of packers of fresh prepared green vegetables in the United Kingdom have been experimenting with O2 mixtures between 70 and 100% (Day 1996). The treatment, referred to as "oxygen shock" or "gas shock," has been found to be very effective in inhibiting enzymatic discoloration, preventing anaerobic fermentation reactions, and inhibiting aerobic and anaerobic microbial growth. High levels of O2 can inhibit the growth of both anaerobic and aerobic microorganisms since the optimal O2 level for growth (21% for aerobes, 0-2% for anaerobes) is surpassed. However, there have also been reports of high O2 (that is, 80-90%) stimulating the growth of foodborne pathogens such as Escherichia coli and Listeria monocytogenes (Amanatidou and others 1999). Recent studies by Kader and Ben-Yehoshua (2000) and Wszelaki and Mitcham (2000) examining the use of superatmospheric O2 levels to control microorganisms on produce, have found that only O2 atmospheres close to 100 kPa or lower pressures (40 kPa O2) in combination with CO2 (15 kPa), are truly effective. These requirements may be difficult to achieve in industry since working with such high O2 levels can be hazardous due to flammability issues. As with most MAP gases, superatmospheric O2 has varied effects depending on the commodity, and further research is required in this area to elucidate the utility of this technique in the fresh-cut produce industry. A high O2 MAP group has been formed in the United Kingdom and includes a number of industry groups, notably Marks and Spencer plc, one of the first retail chains to distribute MAP foods. More recently, the "high O2 MAP club" has provided a base for the new "Novel Gases MAP Club" in the United Kingdom , a group that will investigate the use of novel high O2, argon and nitrous oxide MAP for extending shelf life and quality of fresh-cut produce. Their main focus is research into the commercial application of this process.

Nitrogen has three uses in MAP: displacement of O2 to delay oxidation, retardation of the growth of aerobic spoilage organisms and action as a filler to maintain package conformity (Parry 1993). Of the three major gases used in MAP, CO2 is the only one that has significant and direct antimicrobial activity. A number of theories have been suggested to explain this antimicrobial effect. In general, CO2 in MAP results in an increased lag phase and generation time during the logarithmic phase of growth of the organisms involved (Phillips 1996), with inhibition being concentration and temperature dependent. Theories to explain the antimicrobial action of CO2 have been summarized by Farber (1991):

  • alteration of cell membrane function including effects on nutrient uptake and absorption;
  • direct inhibition of enzymes or decreases in the rate of enzyme reactions;
  • penetration of bacterial membranes leading to intracellular pH changes;
  • direct changes to the physico-chemical properties of proteins.

The inhibitory action of CO2 has differential effects on microorganisms. Thus, while aerobic bacteria such as the pseudomonads are inhibited by moderate to high levels of CO2 (10-20%), microorganisms such as lactic acid bacteria can be stimulated by CO2 (Carlin and others 1989; Amanatidou and others 1999). Furthermore, pathogens such as Clostridium perfringens, C. botulinum and L. monocytogenes are minimally affected by CO2 levels below 50%, and there is concern that by inhibiting spoilage microorganisms, a food product may appear edible while containing high numbers of pathogens that may have multiplied due to a lack of indigenous competition (Farber 1991; Zagory 1995; Phillips 1996). More research needs to be done on the interactions of the background microflora with foodborne pathogens in various modified atmospheres used for produce, as well as on the effects of different gaseous environments on the survival and growth of bacterial foodborne pathogens on whole and fresh-cut produce. The optimal MAP conditions for produce quality and respiration for a number of fruits and vegetables are listed in Table VI-4.

1.1. Types of CAS

There are a number of commercially available systems for CAS (Table VI-1). There are advantages and disadvantages to each, and generally the active control systems are most costly due to the need for constant maintenance of gas levels. Since this type of storage is used for large quantities of product, and is not a type of packaging, it will not be discussed further.

1.2. Types of MAP

MAP is used mainly for wholesale/foodservice use and retail display. There are many MAP systems currently in use (Table VI-3). The films used for MAP of fresh-cut produce (Table VI-2) will be discussed briefly.

1.2.1. Bulk packaging

Bulk packaging is similar to CAS, but relies on a passively modified atmosphere (Lee and others 1996). This type of packaging is mainly used for pallet bags and paperboard containers used in the transportation and storage of commodities. These films can be ecologically advantageous since some are returnable, and thus reuse is possible in some cases. An example of this kind of packaging is the Marcellin System (Table VI-3).

1.2.2. Pre-packaging

Following bulk packaging and arrival at the retail outlet, produce can be pre-packaged by the grocer or the customer using a passive MAP (Smith and Ramaswamy 1996). Pre-packaging in the store usually involves the use of plastic film packaging such as low density polyethylene (LDPE), polyvinychloride (PVC) or polypropylene (PP); films which help to minimize moisture loss and maintain produce quality (Lee and others 1996). However, the films used often supply only a narrow range of gas selectivity and due to its imprecise nature, this type of packaging is only applicable to a few products.

In-store packaging has recently been applied to the new online grocery shopping and delivery services. A number of online "grocery stores," such as and in the United States , and and in Canada , have recently been established. The most comprehensive of these organizations will ship everything from fresh fruits and vegetables and staples to frozen and fresh meat and seafood products. It does not appear that these services use any special packaging in addition to that already used in the grocery store. However, it is recommended that attention be paid to this growing online service, especially in terms of the potential for cross-contamination in the warehouse and temperature abuse during storage and/or transportation.


1.3. Films used in MAP

The use of MAP for whole and fresh-cut produce involves careful selection of the film and package type for each specific product and package size (Tables VI-2 and VI-3). Effective MAP of produce requires consideration of the optimal gas concentration, product respiration rate, gas diffusion through the film, as well as the optimal storage temperature in order to achieve the most benefit for the product and consumer. In addition, when selecting an appropriate film, one has to take into account the protection provided, as well as the strength, sealability and clarity, machineability, ability to label, and the gas gradient formed by the closed film (Zagory 1995).

Recently, the long list of films and commercially available MAP systems has been augmented with the conception of both smart and edible packaging systems (Guilbert and others 1996; Phillips 1996). "Smart" or "intelligent" packaging is being used in the fresh-cut industry and includes indicators of time and temperature, gas composition, seal leakage, and food safety and quality (Rooney 2000). Some intelligent systems alter package oxygen and /or carbon dioxide permeability by sensing and responding to changes in temperature. Other smart films incorporate chemicals into packets placed in the packaging system, with no contact with the product; an example would be the use of O2 scavengers with O2 indicators. Another type of smart film, developed with food safety in mind, is currently undergoing testing. This novel system, when incorporated into a packaging film, uses an antibody detection system to detect pathogens, and expresses a positive finding as a symbol on the surface of the package, thereby alerting food handlers to the presence of pathogens. Although this technology shows promise, it is still in its infancy and comprehensive assessments have yet to be performed. Several limitations have been suggested with this technology; for example, it would not likely be able to detect pathogens at concentrations below 104 CFU/g or cm2 and would not detect pathogens within the product.

Edible biodegradable coatings are yet another variant of the smart film technology, where a film is used as a coating and applied directly on the food (Guilbert and others 1996; Francis and others 1999). Wax has been used in China since the 12th and 13th centuries as an edible coating to retard desiccation of citrus fruits, and in the last 30 years, edible films and coatings made from a variety of compounds have been reported. Guilbert and others (1996) and Baldwin (1994) have extensively reviewed some of the newer edible films (see Tables VI-3 and VI-5). These films are gaining popularity due to both environmental pollution and food safety concerns (Padgett and others 1998). However, a number of problems have also been associated with edible coatings. For example, modification of the internal gas composition of the product due to high CO2 and low O2 can cause problems such as anaerobic fermentation of apples and bananas, rapid weight loss of tomatoes, elevated levels of core flush for apples, rapid decay in cucumbers, and so on (Park and others 1994).

Edible films may consist of four basic materials: lipids, resins, polysaccharides and proteins (Baldwin and others 1995). Plasticizers such as glycerol as well as cross-linking agents, antimicrobials, antioxidants, and texture agents can be added to customize the film for a specific use (Guilbert and others 1996). Plasticizers have the specific effect of increasing water vapor permeability. Therefore, their addition must be considered when calculating the desired water vapor properties of each specific film, since too much moisture can create ideal growth conditions for some foodborne pathogens. The most common plasticizer used to cast edible films is food-grade polyethylene glycol, which is used to reduce film brittleness (Koelsch 1994).

Lipids, or waxes and oils, and resins such as shellac and wood rosin have been widely used for intact fruits and vegetables in two distinct forms, laminates and emulsions (Baldwin and others 1995). Lipid-based edible barriers are known for their low water vapor permeabilities. Koelsch (1994) found that the water vapor permeability of a cellulose-based emulsion barrier is dependent on the lipid moiety used; a minimum permeability can be achieved when stearic acid is used as the lipid. This is due to the effective barrier formed by stearic acid through an interlocking network. However, lipid-based edible films also require a support matrix to reduce brittleness, and have difficulty adhering to the hydrophilic cut surfaces of fruits and vegetables (Koelsch 1994; Baldwin and others 1995). Some of the most common compounds used for support matrices are modified celluloses of hydroxypropylmethyl, ethyl and methylcellulose, chitosan and whey protein isolate (WPI; Koelsch 1994).

In general, polysaccharides such as cellulose, pectin, starch, carrageenan, and chitosan, can adhere to cut surfaces of produce and effectively allow gas transfer; however, they are not effective moisture barriers. Due to their CO2 and O2 permeabilities, polysaccharide-based films allow the creation of desirable modified atmospheres, an attractive advantage over plastic or shrink wrap MAP which can be labor intensive, expensive and environmentally harmful (Baldwin and others 1995). A number of cellulose derived coatings are available commercially, most taking advantage of the modified atmosphere effect of the barriers. Pro-long (Courtaulds Group, London) and Semperfresh (Surface Systems International, Ltd., Oxfordshire, U.K.) are examples of water-soluble composite coatings comprised of the sodium salt of carboxymethyl cellulose (CMC) and sucrose fatty acid ester emulsifiers (Baldwin and others 1995). Their properties are discussed in Table VI-6. A newer product called "Snow-White," based on sucrose esters of fatty acids, has also been used to combat oxidative browning in the potato industry. Nature-Seal is a polysaccharide-based surface treatment that uses cellulose derivatives as film formers, but unlike Semperfresh and Pro-long, does not contain sucrose fatty acid esters. Nature-Seal is a browning inhibitor that is applied as a dip or spray and has been shown to delay ripening of whole fruits and vegetables, and to retard discoloration of peeled carrots and cut mushrooms.

Finally, proteins such as casein, soy, and zein, can also adhere to hydrophilic cut produce surfaces and are easily modified to form films; however, they also allow water diffusion (Baldwin and others 1995). Unlike lipid-based barriers, protein-based barriers do not require the addition of a support matrix, since the protein acts as both the water vapor barrier and structural component of the film (Koelsch 1994). Park and others (1994) reported the successful application of a corn-zein film to extend the shelf life of tomatoes. Color change, loss of firmness, and weight loss during storage were delayed, and shelf life was extended by 6 d in comparison to untreated tomatoes. The corn-zein product used in the above study was a commercial product that was brushed onto the tomatoes (Regular Grade F4000, INC Biomedicals, Inc.), and consisted of 54 g of corn-zein, 14 g of glycerine, and 1 g of citric acid dissolved in 260 g of ethanol. Park and others (1994) did not comment on the use of citric acid in the film solution; however, others have found that edible films composed of zein were more successful in preventing the rancidity of nuts when citric acid was added (Guilbert and others 1996).

In order to obtain an edible film that incorporates all the best qualities of these four basic materials, as well as fulfilling the specific conditions for each fruit or vegetable, manufacturers are now producing films comprised of different combinations. Some of the advantages and disadvantages of the four basic edible film barriers, as well as combinations thereof, are listed in Table VI-5.

As with other MAP technologies, edible films can create a very low O2 environment where anaerobic pathogens such as C. botulinum may thrive; however, antimicrobial compounds can be incorporated into the coating in this scenario (Guilbert and others 1996). Since the antimicrobial or antioxidant can be incorporated and applied directly to the surface of the product, only small quantities are required. Not all films are equally amenable to the addition of antimicrobials. Much of the current work on antimicrobial films is taking place in Europe . Some of the incorporated antimicrobial compounds include metal ions supported in zeolite, isothiocyanate in cyclodextrin with cobalt ion, chitosan, allyl isothiocyanate, silver-based fungicide, quaternary ammonium salt, organic monoglycerides, copper and zinc (Padgett and others 1998), benzoic acid, sodium benzoate, sorbic acid and potassium sorbate and propionic acid (Baldwin and others 1995). Researchers are also currently looking at the use of nisin, a bacteriocin, in coatings to suppress L. monocytogenes, as well as other bacteriocins for the control of C. botulinum. Successful applications of this technology have been demonstrated using sodium caseinate/stearic acid to coat peeled carrots and caseinate/acetylated monoglyceride to coat celery sticks (Guilbert and others 1996). Zhuang and others (1996) investigated the ability of a hydroxypropyl methylcellulose coating containing various antimicrobials to inactivate Salmonella Montevideo on the surface and in the core tissues of tomatoes. Citric or acetic acid (0.2, 0.4%) did not enhance inactivation; however, 0.4% sorbic acid significantly enhanced the inactivation of S. Montevideo , although the tomatoes had a chalky and unappealing appearance. A study performed by Padgett and others (1998) did not specifically look at the application of antimicrobial films to food products; however, the incorporation and behaviour of antimicrobials in edible films were observed. Padgett and others (1998) examined the inhibitory effect of both lysozyme and nisin, incorporated directly into corn zein and soy protein films, against a gram-positive and gram-negative indicator organism. They found that casting, rather than heat pressing during the processing of films, was more effective at producing an antimicrobial film when using corn and soy films. Also, the antimicrobial additives affected the film structure as fracture lines were noted at the microscopic level when lysozyme was incorporated, potentially affecting the film integrity. Following incorporation into the films, both nisin and lysozyme maintained their antimicrobial capacity against the indicator organisms Lactobacillus plantarum and E. coli, which was augmented by the addition of a chelating agent such as EDTA (Padgett and others 1998).

In addition to the study performed by Padgett and others (1998), there have been many studies investigating the migration of additives such as antimicrobials from coatings into food (Guilbert and others 1996). Sodium benzoate, benzoic acid, propionic acid, and potassium sorbate are also generally recognized as safe (GRAS) food additives, and sorbic acid has become a model additive for migration studies. In general, wheat gluten-glycerol films containing lipid components, resulted in a 50% reduction in the diffusivity of sorbic acid out of the film. Films composed entirely of lipids allowed even less diffusion of sorbic acid. Therefore, the most advantageous use of these films for antimicrobial properties would be the formation of a monolayer lipid and sorbic acid film, or a bilayer film composed of a hydrophilic base layer coated with a thin layer of lipid containing sorbic acid (Guilbert and others 1996). Chen and others (1996) attempted the construction of an antimicrobial film containing chitosan (water resistant) and methylcellulose (water susceptible), and either sodium benzoate or potassium sorbate as antimicrobials. Although the film was found to be inhibitory to fungi as judged by inhibitory zones on agar media, release of the antimicrobials from the film was too high to maintain a continuous and effective concentration of the antimicrobial in the film.

Antimicrobial compounds have also been used with traditional films such as low-density polyethylene (LDPE); for example, the fungicide Imazalil (IM) and the antimicrobial grapefruit seed extract (GFSE) have recently been used with bell peppers and lettuce, respectively (Miller and others 1984; Han 2000). In the study using IM, it was noted that the use of IM and IM impregnated film was more effective than either treatment alone at controlling fungal decay; however, IM impregnated film increased the incidence of bacterial soft rot (Miller and others 1984). The action of this fungicide on potential pathogens is unknown. Lee and others (1998) investigated the ability of GFSE with LDPE films to inhibit growth of E. coli, Staphylocuccus aureus, molds, yeasts, and lactic acid bacteria, using the plate disk test. Films containing 1.0% GFSE in LDPE film inhibited E. coli and S. aureus as demonstrated by a clear zone; however, molds, yeasts and lactic acid bacteria were unaffected. After testing the films using the plate disk test, Lee and others (1998) used the films for packaging of curled lettuce and soybean sprouts. Although inhibition of E. coli and S. aureus was not measured on the commodities, it was noted that incorporation of 1.0% GFSE into the LDPE film decreased the growth rates of aerobic bacteria and yeasts (initial counts ranging from 103 to 104 CFU/g) over 8 d for the curled lettuce stored at 5 °C (41°F). Soybean sprouts were found to have a higher initial load of aerobic bacteria and yeasts than the curled lettuce (106 CFU/g). Therefore, the only observed decrease in growth rates was for the lactic acid bacteria over a 12-d period at 5 °C (41°F). Further tests need to be performed using this film technology to ascertain the effects on pathogens as well as aerobic bacteria and yeasts when the film is used with a food product. Inhibition of non-pathogenic organisms that can be indicators of organoleptic quality may lengthen shelf life such that outgrowth of pathogens is possible, while the product is still organoleptically acceptable. Grapefruit seed extract is reported to be inhibitory to a number of human pathogens. There has been evidence, however, that any antibacterial activity of commercial preparations is due to the various preservative agents (triclosan, methyl parabene, benzethonium chloride) contained within the product. Researchers have found that products not containing any preservatives and several self-made preparations had no antimicrobial activity (Woedtke and others 1999). In the aforementioned study by Lee and others (1998), the composition of the GFSE incorporated in the film was not discussed or examined. It is obvious that if pure GFSE is to be used, its antimicrobial properties will have to be fully investigated. If the active antimicrobial ingredients in commercial GFSE preparations are preserving agents, they may be better targets for investigation.

At present, the area of edible films and antimicrobial edible films is not considered a priority by industry due to overall public perception and hesitation about adding more chemicals, natural or not, to fresh produce. Besides the waxing of fruits, edible films are not commonly used and presently, the main issue involves the production of coatings with good surface tension that will stick to produce.

2. Factors affecting shelf life

A main aim of MAP is extension of shelf life. It must be reiterated here that extension of product shelf life may allow outgrowth and/or growth of pathogens to higher levels as compared to air-stored samples. Since fruits and vegetables are still alive and therefore respiring when harvested and processed, there are many factors that affect the post-harvest shelf life extension of fresh produce and the success of MAP.

The rate of respiration of a fruit or vegetable is inversely proportional to the shelf life of the product; a higher rate decreases shelf life (Day 1993; Lee and others 1995). In general, those products with increased wounding, as in the case of fresh-cut produce, will have a high degree of perishability due to increased respiration rates. Respiration can be measured by the oxygen uptake or by production of CO2 and also results in the production of heat and water vapor (Zagory 1995). Therefore, a goal of MAP is to decrease the produce respiration rate, which can be successfully achieved with decreased O2 levels (1-5%) and refrigeration. However, O2 concentrations below 1-2% can lead to anaerobic respiration and the production of off-odors, as well as create ideal conditions for pathogens such as C. botulinum. As previously discussed, high O2 (70-100%) combined with CO2 for MAP has been tested and shown to have beneficial effects on product quality (Amanatidou and others 1999); however, more research is required to support and explain this concept (Wszelaki and Mitcham 2000; Kader and Ben-Yehoshua 2000).

The delay of senescence, the natural form of deterioration, is the main goal in the preservation of fresh produce, as senescence accounts for the majority of post-harvest losses (Lee and others 1995). Senescence is endogenously controlled and is the stage when extensive catabolic reactions occur, resulting in dissolution of plant membranes. It is marked by chlorophyll loss, decreases in RNA and protein content and tissue softening. Plants, for example, senesce to re-route materials into seeds representing the next generation; it is therefore a pre-destined apoptosis process that can only be delayed, not completely inhibited. It is driven by an increase in respiration, as well by an increase in ethylene production in some products, a process referred to as climacteric. It is therefore reasonable to assume that maintaining and reducing ethylene perception and production may effectively delay senescence.

As mentioned above, ethylene, a plant hormone, plays a large role in shelf life and can cause a marked increase in respiration rates and enhance ripening and senescence (Nguyen-the and Carlin 1994; Day 1993). In some commodities, accelerated ageing and the initiation of ripening can occur following exposure to ethylene concentrations as low as 0.1ml/l (Lee and others 1995). As senescence begins, spoilage due to indigenous bacteria can be augmented. Ethylene is also a byproduct of the aerobic combustion of hydrocarbons, and it is therefore important during the handling of produce to maintain low levels of environmental ethylene, which are often increased by fork lifts and other machinery (Zagory 1995). Different biological structures of assorted produce varieties contribute to the product's sensitivity response to ethylene, as well as the response to O2 and CO2. Furthermore, different stages of maturity, cultivar and post-harvest storage conditions also influence sensitivity to ethylene (Lee and others 1995). Control measures taken to minimize perception and production of ethylene following harvest include storage in a modified atmosphere at optimal low temperatures (just above the chilling or freezing injury threshold) and oxidizing the ethylene by various chemical and physical means. Part of the success of MAP, and the quality attributed to MAP products, depends on preventing the damaging effects of exposure to ethylene. To this end, CO2 can inhibit ethylene action as well as autocatalytic production of ethylene by climacteric products such as apples and tomatoes. However, increased damage to whole leaf plants has been observed at CO2 levels above 15-20%, thus reinforcing the importance of designing a specific MAP for each product (Lee and others 1995).

Successful control of both product respiration and ethylene production and perception by MAP can result in a fruit or vegetable product of high organoleptic quality; however, control of these processes is dependent on temperature control. Along the whole food continuum, that is, processing, storage, transportation and retailing, one needs to maintain optimum temperatures. Maintaining proper storage temperatures is often most difficult at the retail level, due to the increased handling and the need to make the product visually appealing. A study by LeBlanc and others (1996) revealed the extent of temperature abuse of produce in the retail setting. Of 746 and 745 produce samples examined during the winter and summer, 87% and 93% of samples, respectively, that should have been stored at 4°C (39.2°F), were being held above 4°C (39.2°F) and as high as 8.4°C (47.1°F). Furthermore, temperature fluctuations between items stored in different parts of the cabinet were observed. The authors stated that MAP products should probably not be stored with fresh fruit and vegetables. For some products, the success and microbiological safety of MAP is dependent on controlled low temperature storage and the product's characteristics. Many MAP fresh-cut products overtly spoil before becoming microbiological safety concerns and thus, the risk factors, that is, outgrowth of pathogens, for both the upper and lower limits of recommended storage temperatures for MAP produce, should be carefully considered when designing a MAP system. Hintlain and Hotchkiss (1987) presented the concept of a safety index where products that result in an increasing ratio of spoilage organisms to pathogenic organisms can be considered less hazardous than products that show a decrease in spoilage organisms with respect to pathogens. This concept could be used when designing MAP systems, with a better understanding of the interaction between spoilage organisms and fresh-cut produce. State of the art temperature control cabinets are currently being used at the retail level; however, it is a matter of recovery on invested capital and managing the system. Recent advances in the cold-storage industry show promise for improved temperature control of produce during transport as well as at the retail level. Freshloc Technologies, Inc. recently revealed a state-of-the-art, wireless, Internet-based data collection system for the transportation of temperature sensitive products. This system automatically monitors and alerts grocery industry personnel to fluctuations in storage temperature and can be adapted to the grocery, restaurant or transport industry. This system should help in maintaining consistent storage temperatures; however, it cannot resolve the problems associated with cabinet design (temperature fluctuations), or the efforts of grocery personnel to make displays as attractive as possible, while neglecting recommended storage temperatures. Mild abuse temperatures will not only shorten product shelf life, but will also allow for the more rapid growth of psychrotrophic pathogens in some products.

3. Influence of MAP/CAS on growth and survival of organisms on produce

3.1. Spoilage organisms

The commonly encountered microflora of fruits and vegetables are Pseudomonas spp., Erwinia herbicola, Flavobacterium, Xanthomonas, Enterobacter agglomerans, lactic acid bacteria such as Leuconostoc mesenteroides and Lactobacillus spp., and molds and yeasts (Nguyen-the and Carlin 1994; Zagory 1999). Although this microflora is largely responsible for the spoilage of fresh produce, it can vary greatly for each product and storage conditions. Temperature can play a large role in determining the outcome of the final microflora found on refrigerated fruits and vegetables, leading to a selection for psychrotrophs and a decrease in the number of mesophilic microorganisms. Previous studies have shown that cabbage (in coleslaw) deteriorated at the same rate at 7°C (44.6°F) and 14°C (57.2°F); however, at 7°C (44.6°F), the reduction in the total microbial load was significant (King and others 1976). Similar phenomena have been reported for shredded chicory salads (Nguyen-the and Prunier 1989) and shredded carrots (Carlin and others 1989), where the total counts of the mesophilic flora decreased with temperature. Low temperature storage not only decreases the growth rate of foodborne pathogens but also increases the inhibitory effects of MAP by increasing the solubility of CO2 in the liquid phase surrounding a food.

The effect of MAP on lactic acid bacteria can vary depending on the type of produce packaged. The increased CO2 and decreased O2 concentrations used in MAP generally favor the growth of lactic acid bacteria. This can expedite the spoilage of produce sensitive to lactic acid bacteria, such as lettuce, chicory leaves and carrots (Nguyen-the and Carlin 1994). The effect of MAP on yeasts is negligible, however, molds are aerobic microorganisms and therefore CO2 can cause growth inhibition at concentrations as low as 10% (Molin 2000), although the effect is not fungicidal (Littlefield and others 1996). Beuchat and Brackett (1990a) examined the effects of cutting, chlorine dip and modified atmospheres on the growth of yeasts and molds on lettuce. At 10°C (50°F), in both air and a modified atmosphere (3% O2 and 97% CO2), and with or without chlorine treatment, the organisms grew slowly regardless of the conditions and at 5°C (41°F), the growth over a 15-d period was erratic. However, no specific inhibitory effects of the modified atmosphere were noted.

The concern when using MAP for fruit and vegetables arises from the potential for foodborne pathogens, which may be resistant to moderate to high levels of CO2 (< 50%), to outgrow spoilage microorganisms, which may be susceptible to the modified atmosphere (Bennik and others 1998). The interaction of the pathogenic and resident (saprophytic) microflora has been extensively reviewed for meat and milk products; however, data are still required for MAP fruits and vegetables (Nguyen-the and Carlin 1994). This interaction of the resident microflora and pathogenic organisms on MAP produce needs to be studied more extensively (Francis and O'Beirne 1998).

3.2. Pathogenic organisms

There are many steps involved along the whole farm to fork produce chain and, therefore, many points for potential microbial contamination (NACMCF 1999). Pre-harvest contamination of fresh produce can occur through the use of non-pasteurized manure for fertilization, fecal contamination by indigenous or domestic animal species as well as agricultural workers, contaminated irrigation water, and general human handling (see Chapter I). During harvest and post-harvest, critical points for contamination include contaminated wash water or ice, human handling, animals, contaminated equipment or transportation vehicles, cross-contamination, and inefficient processing of the product that fails to remove substantial levels of bacteria (NACMCF 1999).

Therefore, MAP produce is vulnerable from a safety standpoint because modified atmospheres may inhibit organisms that usually warn consumers of spoilage, while the growth of pathogens may be encouraged. Also, slow growing pathogens may further increase in numbers due to the extension of shelf life. Currently, there is concern with the psychrotrophic foodborne pathogens such as L. monocytogenes, Yersinia entercolitica and Aeromonas hydrophila, as well as non-proteolytic C. botulinum, although clearly a number of other microorganisms, especially Salmonella spp., E. coli O157:H7 and Shigella spp., can be potential health risks when present on MAP produce.

3.3. Clostridium botulinum

The spores of C. botulinum are commonly found in agricultural soils and on the surfaces of fruits and vegetables. Proteolytic C. botulinum has difficulty growing and producing toxin at temperatures below 12°C (53.6°F), pH below 4.6, a water activity below 0.95 and NaCl concentrations above 10% (Lund and Peck 2000). Non-proteolytic C. botulinum can grow at a minimum of 3°C (37.4°F), pH above 5.0, water activity above 0.97 and NaCl concentrations above 4%. Therefore, there is some concern about the use of MAP with respect to this organism (Zagory 1995). Depending on the product in a MA package, the level of O2 can decrease rapidly if the product is temperature abused and product respiration increases, leaving a highly anaerobic environment ideal for the growth and toxin production of C. botulinum (Francis and others 1999). In a study looking at this potential in lettuce, cabbage, broccoli, carrots, and green beans packaged under vacuum or in air, Larson and others (1997) found that most often the product was grossly spoiled before significant toxin production was detected (Table VI-7). The probability of botulinal toxin being produced before the product was obviously spoiled was less than 1 in 105 in the foods examined using the standard mouse assay for detection of botulinal toxin. Hao and others (1998) found similar results for shredded carrots and green beans packaged under 4 different films allowing for different oxygen transmission rates. Similar results were obtained by Petran and others (1995) for romaine lettuce and shredded cabbage; that is, all toxin-positive samples were grossly spoiled prior to toxin detection. Larson and Johnson (1999) obtained the same results in a similar study when looking at the incidence of botulinal toxin production on artificially inoculated cantaloupe and honeydew. At abusive temperatures, with the exception of UV-treated samples, samples were obviously spoiled, although they were also considered marginally organoleptically unacceptable when toxin was detected. These findings were supported by the results of Hao and others (1998), which showed that packaged lettuce and cabbage became spoiled before becoming toxic. The study by Larson and Johnson (1999) demonstrated the ability of the spoilage flora to protect against pathogen overgrowth. It is likely, however, that product characteristics such as water activity, pH, respiration rate, initial spore levels, and indigenous microflora play a role in the survival and persistence of the pathogen on MAP produce. For example, in 1987, four circus performers in Sarasota , FL became ill with symptoms of botulism after consuming coleslaw prepared from packaged shredded cabbage purchased three weeks earlier in New Orleans (Solomon and others 1990). Researchers suspected that the cabbage had been packaged using MAP and that contaminated cabbage further contaminated the dressing, leading to the recovery of C. botulinum type A toxin and spores from the dressing. A follow-up study done to determine the possibility of C. botulinum surviving on cabbage in MAP was undertaken, and results indicated that only C. botulinum type A grew and produced toxin in the modified environment when stored at room temperature (Solomon and others 1990). Two isolates used in the follow-up study were obtained when the outer leaves of 88 cabbages were surveyed; 12 of them (13.6%) were found to contain toxin type A strains. However, this high incidence of type A spores may have been due to the origin of this particular product and type of soil. For example, Lilly and others (1996) found that only 0.3% (1 of 337) of sampled shredded cabbage obtained from retail suppliers in the United States contained C. botulinum. However, the products tested had all been stored at 4°C (39.2°F), below the minimum for growth of proteolytic C. botulinum.

Growth and toxin production of C. botulinum before obvious product spoilage has also been observed on Agaricus bisporus mushrooms (Sugiyama and Yang 1975) and potato slices (Dignan 1985). As well, Austin and others (1998) performed challenge studies using both nonproteolytic and proteolytic strains of C. botulinum on MAP fresh-cut vegetables and found that samples of butternut squash (5°C [41°F], 21 d) and onion (25°C [77°F], 6 d) appeared organoleptically acceptable when toxin was detected. It was also demonstrated that toxin production by C. botulinum varied with the vegetables tested. Only nonproteolytic strains growing on butternut squash were capable of producing neurotoxin at temperatures as low as 5°C (41°F ) in 21 d, whereas proteolytic strains were able to produce toxin on all vegetables tested (onion, butternut squash, rutabaga, romaine lettuce, stir-fry and mixed salad), except coleslaw at 15°C (59°F) and higher (Austin and others 1998).

A mixture of proteolytic strains were able to produce botulinum neurotoxin on MAP broccoli, stored at 13°C (55.4°F) and 21°C (69.8°F), however the product was obviously spoiled by the time toxin was produced (Hao and others 1999). During a study of uninoculated vacuum packaged minimally processed green bell peppers, Senesi and others (2000) found that after 7 d at refrigeration temperatures, the environment within the package had become anaerobic and high in CO2, stressing the importance of careful selection of a MAP film and initial gas atmosphere.

Fresh mushrooms and tomatoes have also been shown to contain spores of Clostridium spp., and therefore the possibility of botulism associated with these MAP products must not be ignored (Zagory 1995). However, it is thought that the acidic nature of tomatoes (pH <4.6) does not provide ideal growth conditions for C. botulinum. This theory was supported by the results of Hotchkiss and others (1992) who demonstrated that MAP (1.0-2.9% O2) tomatoes stored at 13°C (55.4°F) and 23°C (69.8°F) became toxic only after becoming severely spoiled, well beyond the point of being organoleptically acceptable. The initial concentration of O2 used for high respiring products such as mushrooms can be very important since it will decrease more quickly, resulting in an anaerobic environment. Also, reducing substances of mushrooms can contribute to a lowering of the oxidation-reduction potential to levels suitable for bacterial anaerobic growth. The primary reason for toxin production, however, is the low O2 concentration (Sugiyama and Yang 1975). The standard industry practice of placing holes in the packaging film can discourage the growth of C. botulinum, although the shelf life of mushrooms is reduced.

A comprehensive survey of 1,118 packages of MAP vegetable products from three large U.S. cities found an overall incidence rate of 0.36% of C. botulinum type A spores (Lilly and others 1996). The absence of outbreaks of botulism linked to MAP produce indicates that C. botulinum may be competitively inhibited under the packaging and resident flora conditions of these products. However, more research needs to be done to examine the potential for growth of C. botulinum in a wide variety of MAP produce stored at mildly abusive temperatures such as 7-12°C (44.6-53.6°F) . In addition, other hurdles besides temperature need to be examined to prevent botulinum toxin production.

3.4. Listeria monocytogenes

Recently, concerns about possible pathogen contamination in MAP produce have focused on L. monocytogenes due to its ability to grow at refrigeration temperatures (NACMCF 1999). Numerous researchers have reported that this organism can remain largely unaffected by MAP, while the normal microflora is inhibited (Amatanidou and others 1999; Francis and O'Bierne 1997, 1998). Thus, although MAP produce can remain organoleptically acceptable, L. monocytogenes, with a reduced microflora and, especially if low levels of lactic acid bacteria are present, can grow at low temperatures to potentially harmful levels during the extended storage life of a MAP produce product.

Early studies showed that L. monocytogenes inoculated onto broccoli, asparagus and cauliflower was unaffected by a modified atmosphere of 3% CO2, 18% O2 and 79% N2 for 10 d at 10°C (Berrang and others 1989a). Further studies by Beuchat and Brackett (1990a) clearly demonstrated that L. monocytogenes increased significantly in number on lettuce stored in a modified atmosphere of 3% O2 and 97% N2. Ringlé and others showed that L. monocytogenes did not grow on shredded lettuce stored at either 4°C (39.2°F) or 8°C (46.4°F) when packaged under air using a semi-permeable film, but did grow on the same product when flushed with nitrogen (Francis and O'Beirne 1997). Francis and O'Beirne (1997) also reported that the growth of L. monocytogenes was stimulated by nitrogen flushing at 8°C (46.4°F). In addition, increasing CO2 levels from 10 to 20% has been reported to stimulate the growth of L. monocytogenes in a surface model system (Amanatidou and others 1999).

Challenge studies conducted by Farber and others (1998) focused on commercially available packaged vegetables and salads, as well as vegetables processed to mimic foodservice conditions. The importance of refrigeration was clearly demonstrated as L. monocytogenes population levels remained constant on all fresh-cut, processed and packaged vegetables stored at 4°C (39.2°F), with the exception of butternut squash and carrots on which the levels increased and decreased, respectively. At 10°C (50°F), the growth of L. monocytogenes was supported on all vegetables tested with the exception of chopped carrots, where the population decreased by 2 log units over 9 d. The inhibitory properties of raw, uncooked carrots and carrot juice on the growth of L. monocytogenes have been previously reported (Beuchat and Brackett 1990b). As well, Jacxsens and others (1999) reported a decline in L. monocytogenes on both Brussels sprouts and carrots packaged under a modified atmosphere (2 to 3% O2, 2 to 3% CO2, and 94 to 96% N2) and stored at 7°C (44.6°F),. Under severe temperature abuse conditions (25°C [77°F], for 1-2d) followed by storage at 4 (39.2°F) or 10°C (50°F), inoculated Caesar salad and coleslaw mix supported the growth of L. monocytogenes, although larger increases were noted for the coleslaw mix. These results support those of Beuchat and Brackett (1990a), who found that L. monocytogenes could grow on lettuce following commercial processing procedures such as chlorine treatment, modified atmosphere and shredding. Jacxsens and others (1999) investigated the behavior of L. monocytogenes and Aeromonas spp. on minimally processed vegetables packaged under either a modified atmosphere (that is, 2-3% O2, 2-3% CO2, 94-96% N2) or air, and found that the organoleptic quality of the produce was obviously decreased before pathogen levels increased significantly, and that the growth of the psychrotrophic pathogens was influenced more by the type of vegetable than the atmosphere used. This phenomenon was also observed for L. monocytogenes inoculated on minimally processed green salads; that is, growth was faster on butterhead lettuce than on green endive and was not observed on lamb's lettuce (Carlin and Nguyen 1994). These observations could also be linked to the varied sensitivities of different produce to MAP conditions which can accelerate tissue senescence and death, thus providing nutrients for growth of the pathogens; however, the organoleptic quality of the lettuce was not noted by Carlin and Nguyen (1994). Both carrots and Brussels sprouts demonstrated inhibitory activity towards Listeria (Jacxsens and others 1999). In general, Aeromonas grew faster than L. monocytogenes on shredded chicory endives and shredded iceberg lettuce, consistent with previously reported growth rates on refrigerated vegetables (Anonymous 1996). Castillejo Rodriguez and others (2000) performed a similar study looking at the growth of L. monocytogenes on trimmed fresh green asparagus stored under air. Populations decreased at 2°C (35.6°F) and 4°C (39.2); however, at 8°C (46.4°F), they increased at a rate of 0.038/h. A modified atmosphere developed in the package with an increase in the levels of CO2 (1.63-15.63% at 8°C [46.4°F], 528 h) and a decrease in O2 (18.13-10.35% at 8°C [46.4°F], 528 h). These conditions did not affect the growth of L. monocytogenes, and the authors concluded that these conditions might allow L. monocytogenes to reach potentially hazardous levels during the shelf life of the product. As with other studies (Garcia-Gimeno and others 1996), only temperature had a significant influence on the growth of L. monocytogenes, and storage temperatures below 4°C (39.2°F) were required to maintain the safety of the product.

The effects of competition between the indigenous microflora and pathogens on MAP produce have not been studied extensively. However, in a recent study, Francis and O'Beirne (1998) used a surface model agar system to examine the effects of storage atmosphere on L. monocytogenes and the competing microflora (Pseudomonas fluorescens, P. aeruginosa, Enterobacter cloacae, Enterobacter agglomerans and Leuconostoc citreum). The findings suggested that MAP conditions (5-20% CO2, balance N2 and 3% O2) might increase the growth rate of L. monocytogenes. They did find that at increased CO2 levels (20%), the growth of the lactic acid bacteria increased, inhibiting the growth of the pathogen, possibly due to the production of well-known antilisterial agents. Pseudomonads had no effect on the growth rate of L. monocytogenes in this study, although the fluorescent pseudomonads have been previously shown to activate the growth of L. monocytogenes in various foods, a phenomenon related to the release of potential nutrients by the pseudomonads (Liao and Sapers, 1999; Nguyen-the and Carlin, 1994). Alternatively, Liao and Sapers (1999) also reported that P. fluorescens strains inhibited the growth of L. monocytogenes on endive leaves and spinach, possibly due to the production of a fluorescent siderophore by the pseudomonads. In general, at 3% O2, a level often reached in commercial MAP packages, it appeared that growth of the inoculated mixed natural population was decreased, whereas L. monocytogenes proliferated.

Reports of L. monocytogenes growing on sliced apples in controlled atmosphere (Conway and others 1998) and peeled potatoes in vacuum-packages (Juneja and others 1998) at abusive temperatures provide further evidence that this organism may pose a safety risk with respect to certain MAP fruit and vegetable products, and reiterates the importance of Good Agriculture Practices (GAP), Good Manufacturing Practices (GMP) and HACCP for produce post-harvest handling and processing.

More research needs to be done to examine the influence of different atmospheres, background microflora and storage temperatures on the survival and growth of L. monocytogenes on MAP fresh-cut produce.

3.5. Aeromonas hydrophila

Aeromonas spp. can be found on a wide variety of foods, as well as in most aquatic environments and most often causes gastroenteritis, and occasionally septicemia ( Kirov 1997). Additional information regarding pathogenesis may be found in Chapter IV. Similar to L. monocytogenes, A. hydrophila can grow at refrigeration temperatures, and several studies have shown that growth is not affected by low O2 levels (1.5%) and CO2 levels up to 50% (Francis and others 1999). A survey of 97 prepared salads found A. hydrophila to be present in 21.6% of them, significantly lower than in meat products tested (Fricker and Tompsett 1989). Hudson and De Lacy (1991) also did a small survey of 30 salads and found A. hydrophila in only one salad package not containing mayonnaise. They surmised that the mayonnaise lowered the pH of the food, thereby inhibiting the growth of or inactivating the aeromonads present. Differences in recovery rates between the studies of Fricker and Tompsett (1989) and Hudson and De Lacy (1991) may be due to methodology. Fricker and Tompsett (1989) tested three solid media for recovery of Aeromonas spp. and may have subsequently achieved better recovery, as compared to Hudson and De Lacy (1991) who only used one plating medium. Garcia-Gimeno and others (1996) observed a decline in bacterial numbers paralleled by a decrease in pH and increase in CO2, for A. hydrophila inoculated vegetable salads stored at 15°C (59°F). At 4°C (39.2°F), the bacteria survived, but did not grow. The actual significance of finding A. hydrophila on foods is unclear at the present time.

Berrang and others (1989b) determined that although at both 4°C (39.2°F) and 15°C (59°F), the shelf life of broccoli, asparagus and cauliflower was prolonged by MAP (that is, 11-18% O2, 3-10% CO2, 97% N2), it did not negatively affect the growth of resident or inoculated A. hydrophila. Interestingly, the organism was detected on most lots obtained from the commercial producer. Therefore, for storage periods of 8-21 d, depending on the product, A. hydrophila increased from roughly 104 to 108 or 109 CFU/g, and product that appeared suitable for consumption was heavily contaminated with the pathogen. As with L. monocytogenes, the CO2 levels that were inhibitory to A. hydrophila (that is, >50%) also damaged the product (Bennik and others 1995). As previously discussed, the challenge study performed by Jacxens and others (1999) demonstrated that Aeromonas grew faster than L. monocytogenes on minimally processed vegetables in air and MAP and that a decline in the populations of both organisms was observed on Brussels sprouts . A recent study has proposed the use of a Lactobacillus casei inoculum combined with MAP and chill temperatures to reduce the survival and/or growth of A. hydrophila in ready-to-use vegetables such as fresh-cut lettuce (Vescovo and others 1997). Previous studies have shown that an increase in lactic acid bacteria combined with high levels of CO2 (33%) decreases product pH and, therefore, populations of Aeromonas spp. on vegetable salads (Garcia-Gimeno and others 1996). However, the increased level of CO2 could damage the product.

3.6. Other pathogens of concern with respect to MAP produce

Organisms such as Salmonella, Shigella, E. coli, and various enteric viruses, such as hepatitis A, have been implicated in produce outbreaks, and, therefore, there is concern about their behavior under modified atmosphere conditions (Zagory 1995; Amanatidou and others 1999). A 1986 outbreak of shigellosis was traced back to commercially distributed MAP shredded lettuce; 347 people were affected in two west Texas counties (Davis and others 1988). Fernandez-Escartin and others (1989) tested the ability of three strains of Shigella to grow on the surface of fresh-cut papaya, jicama, and watermelon and reported that populations increased significantly when the inoculated product was left at room temperature for 4-6 h. Shigella is not part of the normal flora associated with produce, but can be passed on as contaminants by infected food handlers and contaminated manure and irrigation water.

More recently, an outbreak of Salmonella Newport was reported in the U.K. , associated with the consumption of ready-to-eat salad vegetables (PHLS 2001). To date, nine human cases have been identified with the isolated strain from the implicated salad vegetables having an identical PFGE pattern to three of the human isolates.

In an agar-based study to investigate the effects of high (80-90%) O2 and moderate (10-20%) CO2 concentrations on foodborne pathogens at 8°C (46.4°F), Amanatidou and others (1999) noted little inhibitory action against a number of pathogens. All pathogens were able to grow in air; however, S. Typhimurium grew slowly, at a rate of 0.011 µ/h. Ten to 20% CO2 was inhibitory to S. Enteritidis; however, S. Typhimurium, L. monocytogenes and non-pathogenic E. coli were unaffected or stimulated. Only when high O2 (90%) and moderate CO2 levels (10-20%) were used, did consistently strong inhibition of S. Enteritidis and E. coli occur. Kakiomenou and others (1998) however, found that S. Enteritidis numbers decreased on both carrots and lettuce when stored under 5% CO2, 5.2% O2 and 89.9% N2. Salmonella Typhimurium and L. monocytogenes actually had an increased growth rate at these concentrations; growth increased from 0.011 and 0.031µ/h to 0.023 and 0.041 µ/h for S. Typhimurium and L. monocytogenes, respectively. In general, E. coli O157:H7, S. Hadar and S. Typhimurium were only inhibited by CO2 levels that caused damage and spoilage to the produce (Piagentini and others 1997; Amanatidou and others 1999; Francis and others 1999). A modified atmosphere of 3% O2 and 97% N2 also had no significant effect on E. coli O157:H7 inoculated onto shredded lettuce, sliced cucumber, and shredded carrot and incubated at 12 and 21°C (21.6 and 69.8°F) (Abdul-Raouf and others 1993). At 5°C (41°F), populations of viable E. coli O157:H7 declined on stored vegetables; however, at 12 and 21°C (53.6 and 69.8°F), populations increased, demonstrating the importance of refrigeration temperatures in maintaining product safety. Richert and others (2000) who, although not studying MAP, reported that E. coli O157:H7 could survive on produce (broccoli, cucumbers and green peppers) stored at 4°C (39.2°F) and proliferate rapidly when stored at 15°C (59°F). In 1993, there were two foodborne outbreaks of enterotoxigenic E. coli (ETEC) linked to carrots in a tabouleh salad served in New Hampshire and to an airline salad on a flight from North Carolina to Rhode Island (CDC 1994). Although these carrots were of U.S. origin, ETEC is a common cause of diarrheal illness in Mexico and developing countries that import fresh product to North America . Research on the behavior of this pathogen on fresh and fresh-cut product, both under MAP and without MAP, seems warranted.

Information on the survival of the enteric pathogens Y. enterocolitica and Campylobacter spp. on MAP produce is extremely limited, and mainly consists of data involving meat products. These organisms can be recovered from animal reservoirs as well as water sources and theoretically, produce may occasionally become contaminated by the application of natural fertilizers, manure, by wild animal feces, or by contaminated surface or irrigation water (Barton and others 1997; Wallace 1997). There is some evidence that mushrooms may be a source of Campylobacter jejuni; it was isolated from 1.5% of retail, polyvinyl chloride film-wrapped, and fresh mushrooms in 1984 (Doyle and Schoeni 1986). It has been suggested that the mushrooms may become contaminated by harvesters. Two separate studies looking at store bought lettuce (Park and Sanders 1992; Little and others 1999) as well as spinach, radish, green onions, parsley and potatoes (Park and Sanders 1992) did not find any evidence of Campylobacter spp., indicating that there may be minimal risk, regardless of the packaging technology used. Interestingly, Park and Sanders (1992) found evidence of gross contamination by Campylobacter spp. of produce at farmers' markets, suggesting that industrial processing may be effective in removing certain pathogens from fresh produce before retail sale. Alternately, contamination may have occurred at the market. A Canadian study detected no Campylobacter in any of 65 unprocessed or 296 fresh-cut and packaged ready-to-use vegetables such as lettuce, carrot, cauliflower, celery, broccoli, or sliced green peppers (Odomeru and others 1997). A more recent study, investigating the survival of C. jejuni on MAP fresh-cut cilantro and lettuce, found that refrigeration temperatures in combination with a modified atmosphere of 2% O2, 18% CO2 and 80% N2 can be favorable for bacteria (Tran and others 2000). Due to the microaerophilic nature of Campylobacter spp., which require 5% O2, 10% CO2 and 85% N2 for optimal growth, the investigators suspected that a low O2 modified atmosphere may provide an environment conducive to survival of the pathogen. Campylobacter jejuni (initial level 106 CFU/g) was able to survive on cilantro, green pepper, and romaine lettuce packaged under normal air, modified atmosphere, and vacuum, for 15 d of storage at 4°C (39.2°F). After 9 d of storage at 4°C (39.2°F) C. jejuni levels decreased to approximately 104 CFU/g on all three vegetables stored under the modified atmosphere. In contrast, C. jejuni decreased to levels of approximately 102 and 103 CFU/g for all vegetables packaged under normal air and vacuum, respectively.

Studies to determine the behaviour of Y. entercolitica on MAP produce have not been published, however data obtained with meat products indicate that 40-50% CO2 has minimal inhibitory effects on its growth (Francis and others 1999).

Enteric viruses such as Norwalk and hepatitis A (HAV) are often the cause of very large foodborne outbreaks, yet none have been linked specifically to MAP produce, and only HAV has been linked to outbreaks involving fresh produce (Beuchat 1996). Of 14 reports of viral gastroenteritis cited by Hedberg and Osterholm, salad was implicated as the vehicle in 36% (Beuchat 1996). One of the HAV outbreaks was linked to the consumption of commercially distributed lettuce that was washed, sliced and bagged before being distributed to restaurants in Kentucky (Rosenblum and others 1990). A total of 202 people were affected, and the investigation suggested that contamination occurred before distribution to the restaurants. Most often these viruses are transmitted by an infected food handler, through the fecal-oral route. In a survival study, lettuce inoculated with HAV was stored in normal air as well as various MAP conditions at both 4°C (39.2°F) and room temperature (Bidawid and others 2001). The highest rate of virus survival on washed lettuce stored at 4°C (39.2°F) for 12 d, was observed for product stored under 70% CO2:30% N2 and 100% CO2, (that is, 83.6 and 71.6%, respectively). Virus survival was significantly lower at room temperature, which was decreased slightly by the addition of CO2 (that is, > 70%), and the lowest virus survival rate (47.5%) was on lettuce stored in a petri dish under air. It should be noted that these CO2 concentrations would be harmful to the product and are not used in retail. These results are consistent with those of Bagdasaryan (1964), who studied the survival of enteroviruses on radishes, tomatoes, and lettuce stored at 6 to 10°C (42.8-50°F) for periods exceeding the normal shelf life of these products, as well as with those of Badawy and others (1985), who studied the survival of rotavirus on lettuce, radishes, and carrots stored at 4°C (39.2°F) and room temperature. In all these studies, the greatest survival rates were observed at refrigeration temperatures. Studies to determine the role of the competitive bacterial microflora on virus survival may be beneficial.

The protozoan parasites Cyclospora cayatenensis, Cryptosporidium parvum and Giardia lamblia have been the etiologic agents in serious foodborne outbreaks involving berries (Herwaldt 2000), apple cider (Millard and others 1994; CDC 1997), and raw sliced vegetables (Mintz and others 1993), respectively. The behavior of these organisms under MAP is not known. However, the increase in incidence of produce-linked outbreaks due to these organisms indicates that research in this area is necessary. Research is needed to examine the behavior of both foodborne viruses and protozoan parasites on MAP produce.

To date, only two MAP produce products, coleslaw mix (Solomon and others 1990) and ready-to- eat salad vegetables (PHLS 2001), have been implicated in foodborne illness outbreaks of botulism and Salmonella Newport , respectively. As well, although it is unclear how the product was packaged, commercially distributed shredded lettuce caused an outbreak of shigellosis in the United States in 1986 (Davis and others 1988). There has been a noticeable increase in the consumption of fresh fruit and vegetables in the last two decades, and more consumers are choosing the less labor-intensive fresh-cut produce. There has been a parallel rise in the number of produce linked foodborne outbreaks, but not linked to fresh-cut produce packaged under MAP. However, vigilance with respect to the safety of these products must be maintained.


5. Research Needs

  • Investigate the antimicrobial effect of superatmospheric O2 in the fresh-cut produce safety.
  • Study the interactions of the background microflora with foodborne pathogens in various modified atmospheres used for produce, as well as the effects of different gaseous environments on the survival and growth of bacterial foodborne pathogens on whole and fresh-cut produce.
  • Examine the potential for growth of C. botulinum in a wide variety of modified atmosphere packaging (MAP) produce stored at mildly abusive temperatures such as 7-12°C. In addition, other hurdles besides temperature need to be examined to prevent botulinum toxin production.
  • Examine the influence of different atmospheres, background microflora, and storage temperatures on the survival and growth of L. monocytogenes on MAP fresh-cut produce.
  • Investigate the behavior of verotoxin-producing E. coli on fresh and fresh-cut product, both under MAP and without MAP.
  • Explore the survival of the enteric pathogens Y. enterocolitica and Campylobacter spp. and the behavior of foodborne viruses and protozoan parasites on MAP produce.
  • Hurdle technology or the combination of novel methods of food treatment and packaging need to be examined, for example, irradiation used with MAP and antimicrobial films used in combination with MAP.
  • Evaluate the use of intelligent packaging systems.


  Read the full document:


  More very recent 2007 information about Listeria in low oxygen MA packs:




© Copyright 2004 - PerfoTec B.V.