One of the main concerns of the seafood industry is the health risk associated with traditional consumption of oysters, as most bivalves are eaten raw or minimally cooked, and whole, including the viscera. Besides being highly perishable, oysters can pose hazards to public health because their filter feeding behavior leads to the accumulation of pathogens naturally present in growing areas contaminated with polluted water, making the bivalves a high-risk food group (Pardío-Sedas, 2015). To improve their microbiological safety, raw mollusks are often depurated by placing them in controlled, sanitized seawater allowing them to purge contaminants from their tissues by means of their natural pumping activity. The effectiveness of the depuration process depends on the diversity and physiology of the particular mollusk species as well as on water and system characteristics (Schneider et al., 2009). Depuration systems consist of flow-through, or closed, recirculation with chemically (chlorine and/or ozone) or physically (UV irradiation) disinfected water to eliminate bacteria and other pathogens and spoilage microorganisms. The advantage of closed systems is that they use considerably less water than open systems.

Ozone has been given GRAS (Generally Recognized as Safe) approval by the USDA and the FDA for direct contact with food products (United States Food and Drug Administration [USFDA], 2016). Ozone acts on bacteria by oxidation of membrane glycoproteins and/or glycolipids, inhibition of membrane-bound enzymes, and it may also damage DNA due to oxidation of double bonds by singlet oxygen (O’Donnell et al., 2012). Depuration with ozonated water is a postharvest process for shellfish moderately contaminated with fecal coliform bacteria so as to increase the availability and supply of safe, nutritious bivalves such as oysters. However, ozone-depuration efficiency of mollusks with high initial levels of Escherichia coli and Vibrio spp. could lead to insufficient reduction of pathogen loads. Since different microorganisms respond differently to the depuration process, it may fail to guarantee consumer safety (Croci et al., 2002; Food and Agriculture Organization [FAO], 2008). A growing body of research has identified direct application of ozone as a beneficial technology, defined as direct application of residual ozone and ozone-produced oxidants to farmed species of finfish and shellfish species inside the recirculating systems, and is thus distinct from conventional discrete ozone usage. This approach appears to be increasingly employed due to proven enhancement of hygiene, water quality and production, provided dosages are appropriate to maintain animal health and welfare. Appropriate ozone dosages varied from 200 mV up to 600 mV (Powell and Scolding, 2018). The chemistry of ozone in seawater is considerably different from that in brackish and freshwater. When seawater is ozonated, the ozone reacts with a variety of naturally occurring chemical species. The most important reaction is the oxidation of bromide ions (Br^-), forming hypobromite ions (OBr^-) (Oemcke and van Leeuwen, 1998). In seawater with a typical pH of 8, hypobromous acid (HOBr) will predominate and be the most important disinfectant with a half-life of hours to days depending on light conditions and water quality characteristics. Thus, continuous ozonation is beneficial because seawater quality for depuration remains relatively stable (Summerfelt et al., 2009). During depuration, specific water parameters such as pH, temperature and salinity, need to be applied for each species, as changes in these parameters may cause shellfish to reduce or stop activity, thus reducing the effectiveness of the depuration process (Lee et al., 2008). Even though ozone may become less soluble and less stable as temperature and pH increase during depuration, the resulting radical species contribute to its efficacy.

Although the depuration process appeared to be a promising postharvest treatment for minimizing risks of infections associated with raw oyster consumption, it would be necessary to use this in conjunction with other inactivation treatments to achieve better decontamination efficacy. Combined post-harvest techniques have been also studied, such as heat/cool pasteurization (Andrews et al., 2000), high hydrostatic pressure (Ahmadi et al., 2015), and rapid freezing with frozen storage (Liu et al., 2009). However, bivalves die during these processes. Ozone treatment has been proven effective for slowing down the reproduction of spoilage bacteria in mussels, especially when combined with refrigeration (Manousaridis et al., 2005). Nevertheless, efforts to minimize foodborne illnesses through proper refrigeration of postharvest oysters may be compromised as Vibrio cholerae strains exhibit significant differences in survival rates in artificially contaminated seafood at refrigeration or freezing temperatures (Johnston and Brown, 2002). Among the different refrigeration preservation techniques, superchilling technology is one of the most efficient and promising technologies for storing raw seafood. Superchilled food products are stored 1–2°C below the initial freezing point of the product, just below -1.0°C under closely controlled conditions to extend the seafood shelf-life at least 1.4∼4 times that of traditional chilling (Magnussen et al., 2008). In the range of superchilling temperature (0 to -4°C), microbial activity is inhibited, and most bacteria are unable to grow. Accordingly, reduction in microorganism growth rate is due to a synergistic effect of the reduction in water activity and temperature (Kaale et al., 2011). In order to prolong the shelf life of aquatic foods, the effect of superchilling in conjunction with other techniques has been studied, promising preservative potentials in reducing microbiological contents. Sequential application of certain set of preservative factors or hurdles has been shown to result in a higher level of inactivation than the sum of inactivation levels achieved when each preservative is applied separately. This enhanced inactivation is often referred to as a synergism. Synergism is beneficial since the disinfectant dose and the reaction time required for the same level of inactivation can be reduced. A synergistic effect could be achieved if the hurdles in a food hit different targets (e.g., cell membrane, DNA, enzyme systems, etc.) within the microbial cells and thus disturb the homeostasis of the microorganism present (Leistner, 2000). In this regard, refrigeration can be combined with the use of ozone, therefore harnessing the results by a synergic effect (Gonçalves, 2009).

A few studies have reported that refrigeration or superchilling combined with ozonized water have a remarkable effect in reducing the microbial levels in seafood. Manousaridis et al. (2005) reported a shelf-life of 11–12 days for shucked vacuum-packed and refrigerated mussels (4 ± 0.5°C) ozonated 90 min in an ozone-saturated aqueous solution (1 mg/L). Our previous assays carried out during the windy season indicated that direct ozone depuration (0.2–0.6 mg/L) significantly reduced low levels of E. coli and the isolation of V. cholerae in oysters during the superchilled storage period. Biochemical analysis at 0.4 mg O₃/L depuration, indicated no significant (P > 0.05) changes in the LYS levels of oyster meat after 6 h depuration. However, panelists detected a significant (P < 0.05) decrease in firmness, but not in elasticity (Pardío et al., 2010). Likewise, TBA values of 0.4 mg O₃/L depurated oyster samples (2.629 MA/kg) were slightly higher (P < 0.05) than control samples (2.244 MA/kg), but no unpleasant flavors were detected (Rivas Mozo, 2010). Rey et al. (2012) reported that storage of oyster (Ostrea edulis) in ozonated slurry ice (0.2 mg/L ozone) at 0°C ± 2°C provided better control of Enterobacteriaceae than non-ozonated flake ice. Jianbing et al. (2013) reported that superchilling (-1.2°C) combined with ozonized water (1.8 mg/L) decreased the aerobic plate count of pomfret fillets. Recently, Bono et al. (2017) found that superchilled storage at -1°C improved the antimicrobial activity of 0.3 mg/L-ozonized slurry-ice during European anchovy (Engraulis encrasicolus) and sardine (Sardina pilchardus) postharvest preservation. However, to our knowledge, no information is available regarding the effect of ozone depuration combined with superchilled storage on E. coli, Salmonella spp. and V. cholerae survival in bivalves.

In Mexico, oysters are harvested extensively within the oyster-producing areas found along the Mexican Gulf coast. In Mexico, the state of Veracruz is the primary oyster producer, harvesting 26,713 tons annually, which accounts for 43% of the national average annual production (61,996 t) (Comisión Nacional de Acuacultura y Pesca [CONAPESCA], 2017). The American oyster (Crassostrea virginica) is one of the most popular bivalve mollusks, widely consumed in large quantities. They are sold alive in whole shell, shucked in fresh form or packaged and refrigerated in polyethylene bags. Although previous studies have revealed a high prevalence of E. coli, V. cholerae O1/O139 and V. cholerae non-O1/non-O139 chxA in oysters (C. virginica) harvested from estuarine lagoons in Veracruz (Pardío et al., 2010; López et al., 2015), a relatively high proportion of Mexican oysters sold in restaurants and markets is not currently subjected to any post-harvest process. Moreover, along the traid chain these specimens are not maintained in refrigerated conditions and are thus a health hazard. According to the Mexican Norm NOM-242-SSA1-2009 (Secretaría de Salud [SSA], 2009), that provides guidelines for the sanitary control and commerce of shellfish in Mexico, shellstock oysters should be kept alive and adequately refrigerated to an internal body temperature of 7°C and stored 7 days at most to ensure safe consumption. Because of the importance of raw oysters in gastronomy and economics, the improvement of their microbial safety is of major interest. Given the limited data available regarding the microbiological safety of superchilled ozonated shellstock oysters, the aim of this study was to determine the antimicrobial efficacy of superchilled storage on V. cholerae, fecal coliforms, E. coli and Salmonella spp. loads in direct ozone-depurated shellstock American raw oysters and compared with superchilled alone, as a cost-effective postharvest process for effectively reducing contamination without impairment of oyster viability, while ensuring public health.

The presence of fecal coliforms, E. coli, and Salmonella spp. in oysters is indicative of contamination of fecal origin. They can enter the aquatic environment in oyster-growing areas near the densely populated coast where agricultural and industrial activity is intense. The presence of Salmonella bacteria is an important public health issue in Mexico, with high rates of infection reported annually. In 2017, a total of 41,917 cases in Mexico were reported of which 6,116 were in the state of Veracruz (Secretaría de Salud [SSA], 2017). High levels of Salmonella (3.56 log CFU/g) like those obtained in our study were reported in naturally polluted oysters (Saccostrea cucullata) in India (Jana et al., 2013). India has oceanographic conditions similar to those of Veracruz with tropical and warm seawater areas. Salmonella spp. are commonly isolated from seawater in tropical regions and high counts are related to long periods of torrential rainfall, when contamination is transported from source points to the sea via river water (Simental and Martínez-Urtaza, 2008). The Jamapa, Huatusco, Cotaxtla, and Totolapan river plumes that contribute to MLS and water temperature in this tropical area might explain the presence of these pathogens in oysters harvested from this lagoon. Nevertheless, this area is economically important for seafood production and consumption and recreation and the MLS is one of the largest shellfish-producing estuarine lagoons on Mexican Gulf Coast producing and harvesting oysters year-round. Oysters from the MLS supply seafood restaurants and oyster outlets in nearby cities such as Veracruz and Boca del Río. They are also shipped to Cancun, Monterrey, and Mexico City. However, E. coli (600 MPN/100 g) and V. cholerae non-O1/non-O139 (3.0–28.0 MPN/g) have been isolated in oyster samples from the MLS (Pardío et al., 2010; López et al., 2015). From a public health perspective, consumption of these raw oysters should be considered a potential health hazard.

After 6 h ozone depuration the levels were below the maximum tolerable limits set by Mexican regulations for fecal coliforms (400 MPN/100 g) and E. coli (230 MPN/100 g) (NOM-242-SSA1-2009; Secretaría de Salud [SSA], 2009). The Limits for Verification of Depuration Plant Performance from the USFDA National Shellfish Sanitation Program (NSSP) Guide for Fecal Coliform of 20 MPN/100 g (70 MPN/100 g 90th percentiles) (United States Food and Drug Administration [USFDA] and National Shellfish Sanitation Program [NSSP], 2013) were met as well. According to our findings, it is evident that bacteria were not cleared at the same rate. The mechanisms involved in microbial inactivation by ozone are complex and the susceptibility of bacteria to ozone varies among genera, species, and stage of cellular growth. Moreover, microorganisms are inactivated by ozone at different rates, possibly because of differential membrane permeability (O’Donnell et al., 2012). Several studies have reported that fecal coliforms showed considerably less sensitivity to ozone treatments, suggesting that they are more ozone-resistant than the other organisms (Adams et al., 1989; da Silva et al., 1998). Prabakaran et al. (2012) reported that E. coli revealed high sensitivity to ozone (0.1 mg/L for 15 min) compared to Klebsiella pneumoniae. Thus, the effectiveness of ozone in killing coliforms depends on the specific coliform bacteria. Effectiveness may also depend on other factors, such as locations where bacteria are attached, since ozone mainly reduces surface contamination, and inherent differences in bacterial cell envelope, which is a primary target of ozone activity (O’Donnell et al., 2012).

The depuration results showed that there is a difference between the abilities of Salmonella spp. and E. coli to survive in oysters. E. coli concentration in the control oyster sample was reduced 63% while Salmonella spp. decreased 10% after 6 h-depuration. Morrison et al. (2012) found that 15 days after exposure to closed slow flux ASW depuration, an average of 10³ CFU/g of Salmonella Newport LAJ160311 remained in the oyster meat (Crassostrea gigas), while an average of 5.0 CFU/g of E. coli ATCC 25922 survived. Relatively few studies focused on quantification of Salmonella spp. in ozonated and naturally contaminated oysters or bivalves are available. Çolakolu et al. (2014) reported mean initial Salmonella typhimurium loads (8.8 ± 0.4 and 6.4 ± 0.6 log₁₀ CFU/g) in Donax trunculus and Tapes decussatus, respectively. Salmonella typhimurium was the first strain excreted from both clam species after 66 h of ozone depuration with an open-circuit (6 L/min) at 50 mg/h O₃. However, little is currently known about how Salmonella interacts with marine invertebrates. According to Cox et al. (2016), the ssrB regulated effector could contribute to the fitness of Salmonella in oysters through regulatory mechanisms which are not currently understood. ssrB is required for intracellular survival in vertebrate macrophages. Komanapalli and Lau (1998) studied the effect of short-term ozone exposure at 600 mg/L (1–30 min) on E. coli K-12 and observed progressive degradation of intracellular proteins and membrane permeability, cell viability was unaffected, but progressively decreased with longer exposure. In contrast, a Salmonella Enteritidis population in distilled water decreased 6 log at 1.5 mg/L of ozone, as ozone treatment disrupted the cell membranes followed by the lysis reaction affecting cell viability (Dave, 1999). In addition to the damage to microbial cell envelopes, ozone may induce mutagenic effects on Salmonella typhimurium, leading to cell injury or inactivation (Dillon et al., 1992).

The oyster samples reached the legal limit set by Mexican regulations for V. cholerae non-O1/non-139 and Salmonella spp. of absence per 50 g of oyster flesh after 4 and 6 h-depuration with 0.4 and 0.6 mg/L treatments, respectively. Although the study was conducted during the dry season, when the possibility of high levels of contamination increases, results confirm the effectiveness of the depuration system design. Nevertheless, fecal coliforms and E. coli counts did not decrease to zero levels. It has been reported that 0.2% of the originally accumulated bacterial level would likely remain in the shellfish through a 50-h depuration period (Bella et al., 2000). Oysters released E. coli, which reached the legal bacteriological limits after 2 h depuration with 0.4 and 0.6 mg/L treatments, while V. cholerae non-O1/non-139 limits were reached after 2 h and 4 h depuration with 0.4 and 0.6 mg/L, respectively, indicating that E. coli was more sensitive than V. cholerae non-O1/non-139. Several ozone depuration studies have reported that naturally occurring bacteria such as Vibrio spp. are not likely to be removed effectively due to differential reduction rates of bacteria in depurating shellfish. Croci et al. (2002) reported that experimentally contaminated mussels (Mytilus galloprovincialis) showed a reduction of E. coli (42%) after 5 h, while V. cholerae O1 and V. parahaemolyticus declined by 1 log after 24 h depuration with 50 mg/h ozone, remaining almost constant for 44 h. Likewise, Meloni et al. (2008) showed a decrease in E. coli counts after 8 h depuration, while naturally occurring Vibrios declined at a slower rate (6.7%). Çolakolu et al. (2014) reported that E. coli and Salmonella enterica subsp. enterica serovar typhimurium were eliminated in 78 and 66 h, respectively, whereas V. parahaemolyticus was present after 72 h in clams (Donax trunculus and Tapes decussatus), depurated with 50 mg/h-ozone. However, an increase in Vibrionaceae species (21%) was observed in Pacific oysters (Crassostrea gigas) submerged in ozonated water (0.005 mg/L-2 min) (Rong et al., 2010a). These reports differ from our findings. The results presented here indicate the efficacy of 0.4 mg/L- ozone depuration in reducing 93.3% of fecal coliforms and 94.7% E. coli (Table 1), 100% of V. cholerae non-O1/non-O139 and Salmonella spp. loads (Table 2), in oysters after only 6 h. When depuration periods above 48 h are implemented, bivalve quality may decrease due to the lack of feed during depuration, causing significant economic losses to stakeholders. As depuration may require higher investment and operation costs, oysters should not be killed during processing but kept alive to retain their biochemical and sensorial quality. Thus, it is critical to determine optimal conditions on a species-by-species basis for shellfish microbial depuration to prevent constraints in shell opening. Considering the disinfection efficiency achieved with the depuration system designed for this study, which implied direct application of residual ozone and ozone-produced oxidants to oyster, this approach appears to be a beneficial technology.

The differences observed in the effectiveness of ozone depuration as compared with other reports may also be due to several factors, such as bivalve species, bivalve-species-specific binding potential of microorganisms due to differing mechanisms of persistence, and sensitivity or relative resistance of microorganisms to ozone. Regarding the sensitivity of microorganisms to ozone, free radical activity of molecular ozone can decrease growth (up to inactivation) of specific microbiological species such as Escherichia coli, Salmonella typhimurium, Staphylococcus aureus, Vibrio parahaemolyticus, and Vibrio cholerae (O’Donnell et al., 2012). Farooq and Akhlaque (1983) observed that lethal concentration of 0.23–0.26 mg/L ozone applied 1.67 min at pH 7.0 and 24°C inactivates E. coli and Salmonella typhimurium suspended in water, while 0.48–0.84 mg/L of ozone applied 15 min to spring water was needed to reduce V. cholerae 95% (Orta et al., 1998). Nevertheless, high growth rates of V. cholerae have been observed in seawater treated with high ozone doses (700 mV) (Hess-Erga et al., 2008). Thus, the effectiveness of ozone treatment depends on ozone concentration, length of ozone exposure (contact time), pathogen species and loads, and levels of organic matter (Lee et al., 2008). Several studies have reported that it takes longer to depurate the Eastern oyster (C. virginica) of Vibrio spp. than of E. coli (Murphree and Tamplin, 1991). The lack of depuration efficiency in reducing Vibrio levels might be due to the adhesion of Vibrio spp. to bivalve’ tissues (Pruzzo et al., 2005). According to these authors, different interactions between soluble hemolymph components and the signaling pathways of the hemocyte bivalve host may be responsible for the persistence of Vibrio species within bivalve tissues. This ability of bivalves to exert control over the microbial community in terms of both abundance and biodiversity should be considered, particularly during depuration. The depurating process represents a metabolic effort/fitness that could compromise bivalve’ viability, especially when the suspended food particles are scarce or under stressful conditions (Antunes et al., 2010). Moreover, this ability may explain why artificially contaminated mollusks depurate more rapidly than environmentally contaminated ones (Jones et al., 1991; Schneider et al., 2009). Generally, bacteria are rapidly reduced in shellfish by using seawater only. In this case, the effectiveness of the depuration process depends on the diversity and physiology of the particular mollusk species. Contaminant reduction in the digestive gland is primarily a function of defecation or digestion or both. Several studies (Kueh, 1987; Doré and Lees, 1995) have demonstrated that bacterial reductions may be predominantly influenced by digestive processes, as gut transit times for mollusks are normally rapid. As a prerequisite for decline in coliform number, oysters must pump water through the mantle cavity (Bayne et al., 1976; Gosling, 2003). In this study, contamination levels of raw oysters decreased after 4–6 h of adaptation period due to oyster physiology as microorganisms that were accumulated during previous feeding activities in the alimentary digestive tract of the animal are discharged as part of the fecal material. It is possible that the rates of reduction observed in this study are the actual times required for Crassostrea virginica to eliminate fecal coliform and E. coli from the digestive tract under depuration conditions and characteristics of the system.

The faster reduction of E. coli could induce producers to adopt shorter depuration times, even though high V. cholerae and Salmonella spp. loads would be present, making the bivalve mollusks a high-risk food and a potential health hazard for consumers. Hence, fecal bacteria and E. coli do not adequately predict the comparatively lower rates and levels of V. cholerae during oyster depuration or the effectiveness of ozone depuration. This ability of oysters to selectively retain some microbial species has profound public health implications, as depuration practices have been strongly focused on eliminating fecal coliforms. E. coli is still used as an indicator of the sanitary quality of bivalve mollusks and their growing areas in México. Therefore, the shellfish industry will be compelled to rely on more efficient post-harvest preservation processes to maintain a commercially viable raw product and protect consumer health, with no significant adverse effects on oyster quality. In recent years, alternative fecal indicators such as fecal anaerobes (genera Bacteroides and Bifidobacterium, spore-forming Clostridium perfringens), viruses [B. fragilis phage, coliphages (F-RNA phage)], fecal organic compounds (coprostanol), and nanobiosensors have been increasingly applied (Savichtcheva and Okabe, 2006). It seems that the use of alternative indicators along with conventional fecal markers is promising to identify fecal pollution and associated pathogens.

According to Tables 3, 4, the combination of ozone and superchilled storage significantly controlled the levels of fecal coliforms, reduced the E. coli and V. cholerae non-O1/non-O139 loads and eliminated Salmonella spp. in both treatments throughout the storage time, compared with control oysters. The observed levels were below the Mexican legal limits. The observation that fecal coliforms and E. coli population differed during storage could be explained by temperature-induced differences in adaptation and competitiveness within this group of spoilage organisms. The abrupt fecal coliform and E. coli mortality on day 5 of storage in control oysters could be linked to the drop-in temperature. These results agree with Hood et al. (1983), who reported that the mean initial fecal coliform count (7.0 × 10⁵ MPN/g) in Crassostrea virginica stored at 2°C decreased to 1.7 × 10⁵ on day 7 but increased to 2.5 × 10⁶ MPN/g on day 14, while E. coli counts in fresh oysters (2.9 × 10³ MPN/g) increased to 2.2 × 10⁴ on day 7 but decreased to 3.0 × 10³ MPN/g on day 14. Several fecal coliform strains from environmental sources are psychrotrophic with a minimum growth temperature range of -5 to +5°C (Leclerc et al., 2001). These findings suggest that fecal coliform tolerance and adaptative survival in cold temperatures may be due to the expression of cold-adaptive proteins other than the previously documented major cold shock proteins such as CS7.4 and CsdA reported in E. coli (Jones et al., 1992), indicating that non-E. coli fecal coliforms survive, responding more efficiently to the superchilling storage temperature. These more cold-adapted species, strongly promoted under refrigeration, can reproduce faster in oysters. However, the exact functioning of this cold-adaptive response in several non-E. coli fecal coliforms remains to be elucidated. In contrast, the analysis of the ribosomal fraction of E. coli cells shifted from 37°C to temperatures below 5°C reveals that, during cold-shock, ribosome profiles undergo a severe reduction in polysomes with a concomitant increase in monosomes and cessation of bacterial growth (Hébraud and Potier, 1999). Moussa et al. (2008) reported that the viability of E. coli cells decreased from 87% after 10 min to 4% after 71 days at -10°C. The loss of cell viability was attributed to exposure to cold shock which induced membrane damage. Thus, cells could be inactivated by the action of sub-zero temperatures alone.

The combination of ozone depuration and superchilling (-1°C) may have decreased the V. cholerae non-O1/non-O139 loads below detection levels. However, the significant increase observed on days 9 and 14 may be related to low-temperature adaptation due to differentially expressed genes involved in the cold shock response of V. cholerae. Townsley et al. (2016) reported that the cold shock gene cspV of V. cholerae is upregulared > 50 upon a low temperature shift from 37 to 15°C. Many of the changes in gene expression are presumably oriented toward overcoming the challenges imposed by cold shock to survive at low temperature. The legal limit of absence in 50 g of oyster meat required by Mexican regulations for V. cholerae non-O1/non-O139 was accomplished on day 9 of superchilled storage with 0.4 mg/L treatment, and during 14 days for Salmonella spp. in both treatments. Our results differed from those reported by Rong et al. (2010b) who observed that the proportion of Vibrionaceae spp. decreased from 20 to 2% on day 5 and was not detected after 60 days in raw non-ozonated Pacific oysters (C. gigas) stored at -3°C. V. cholerae cells in stationary phase can remain viable for long periods and are even capable of producing disease, as bacteria may retain pathogenicity genes/factors (Pruzzo et al., 2003). Moreover, variability of survivability has been observed among bacterial species. In a study of the effects of freezing on survival of Salmonella derby, S. typhimurium, and E. coli in Pacific oysters, both species of Salmonella were highly sensitive to freezing, and 1% or less survived after 48 h regardless of freezing methods (chest freezer at -23°C, freezer at -34°C, and walk-in freezer at -17.8 °C). E. coli was less sensitive to freezing, with a 10–30% survival rate in the oysters after 1 week of storage at -34°C (Digirolamo et al., 1970). In contrast, Gopalakrishna and Shrivastava (1989) reported that Salmonella paratyphi B, the most resistant serotype, survived up to 9 months during storage at -20°C. The accumulated evidence suggests that V. cholerae, E. coli and Salmonella are capable of responding to low temperature stress through production of cold-shock proteins (CSPs) to be cold-adapted to function properly at low temperatures (Jeffreys et al., 1998; Bollman et al., 2001; Phadtare, 2004; Oliver, 2010). Enhanced stress resistance in response to cold shock should draw our attention to the increased risk presented by these pathogens in the seafood industry as low storage temperature is one of the most important processes for controlling the safety of seafood.

A few studies have reported that chilled or superchilled storage combined with ozonated water have been capable of reducing E. coli, V. cholerae and Salmonella spp. in bivalves over time. Manousaridis et al. (2005) reported a shelf-life of 11–12 days for shucked vacuum-packed and refrigerated mussels (4 ± 0.5°C) ozonated 90 min in an ozone-saturated aqueous solution (1 mg/L) as compared with 8–9 days shelf-life for the non-ozonated sample based on microbiological analyses. Our previous assays carried out during the windy season indicated that ozone depuration (0.2–0.6 mg/L) significantly reduced the low levels of E. coli and the isolation of V. cholerae in oysters during the superchilled storage period (Pardío et al., 2010). Rey et al. (2012) reported that storage of oyster (Ostrea edulis) in ozonated slurry ice (0.2 mg/L ozone) at 0°C ± 2°C provided better control of Enterobacteriaceae, with significant (P < 0.05) differences being observed between ozonated slurry ice oysters which levels decreased from 1.63 CFU/g on day 1 to 0.90 CFU/g on day 6 of storage and non-ozonated flake ice which levels increased from 1.24 CFU/g to 2.39 CFU/g on day 6 of storage. Ozone- superchilled storage applied to fishery products has been studied as well. Jianbing et al. (2013) reported that superchilling (-1.2°C) combined with ozonated water (1.8 mg/L) had a remarkable effect reducing the aerobic plate count of pomfret fillets. Recently, Bono et al. (2017) found that superchilled storage at -1°C improved the antimicrobial activity of 0.3 mg/L-ozonized slurry-ice during European anchovy (Engraulis encrasicolus) and sardine (Sardina pilchardus) postharvest preservation, reducing Enterobacteriaceae at below 1 log CFU/g throughout superchilled storage.

According to Table 5, superchilled storage at -1°C significantly decreased the levels of fecal coliforms on day 5 and those of E. coli on day 1. As previously mentioned, fecal coliform tolerance and adaptative survival in cold temperatures have been observed. The initial level of Vibrio cholerae non-O1/non-O139 in non-ozonated oysters on day 5 (175.0 MPN/g) of superchilled storage decreased 78.3% (38.0 MPN/g). These results are similar from those reported by Rong et al. (2010b) who observed that the proportion of Vibrionaceae spp. decreased from 20 to 2% (90%) on day 5 in raw Pacific oysters (Crassostrea gigas) stored at -3°C. Meanwhile, Salmonella spp. counts decreased 35.6% from 3.845 to 1.37 10³CFU/g. Gopalakrishnan et al. (2016) isolated Salmonella spp. and Vibrio cholerae from the various parts of Rastrelliger kanagurta (Indian mackerel) stored at -5°C for 5 days, with a microbial load of 6.1 × 10⁴ CFU/g. Thus, given the antimicrobial effect observed in ozonated oysters, there is a beneficial effect of combining ozone depuration and superchilled storage at -1°C on the microbiological safety of oysters. Our results suggest an additive effect of superchilling and ozone pre-treatment in altering microbial integrity and survival on oysters, although some strains may be intrinsically more resistant to cold shock injury than other strains. Several studies have indicated that the residual ozone concentration is greatest at low temperature (O’Donnell et al., 2012). In this context, superchilling storage may facilitate the ozone decomposition mechanism, improving oxidation of membrane glycoproteins and/or glycolipids and the inactivation of microbial cells. From this perspective, controlling the growth of these bacteria through ozonation may be important to improve preservation of oysters during superchilled storage. Improvements in the microbial safety of oysters can have an important economic impact by reducing losses, improving marketability and ensuring public health. In this regard, ozone depuration combined with superchilled storage improved the microbial safety of oysters by a synergic effect.

The mortality rate observed in this study was lower than the 52.5% reported by Maziero and Montanhini (2015) for mangrove oysters (Crassostrea brasiliana) after 15 days of storage at 5°C. Although no reports of mortality percentage for ozonated oysters stored at superchilled temperatures were found, it has been reported that eastern oyster Crassostrea virginica, an eurythermal suspension-feeding bivalve, is able to tolerate a minimum temperature of -2°C (Pernet et al., 2008). During long-term air exposure at cold temperatures oysters can depress metabolic rate and respire anaerobically and are able to survive without available oxygen and seawater. A survival rate of 35.7% after 50 days of air exposure at 4°C has been reported, and the accumulation of acidic end-products and a decline in pH (6.92 to 6.65) were observed within 2–3 days (Kawabe et al., 2010). The decrease in pH of the live bivalves represents the beginning of deteriorative processes that accelerate postmortem. It has been reported that oysters (with liquor) were classified as being of good quality if their pH ≥ 6.0 (Schneider et al., 2009). Intracellular pH is one of the important factors controlling metabolic rate. Moreover, changes in pH are related to shell movement. During shell closure, pH decreases gradually, suggesting an initial reliance on anaerobic metabolic pathways to sustain life (Kawabe et al., 2010). Demonstration of stable pH during storage indicates maintenance of extracellular pH to maintain respiratory gas exchange, but when death occurs the pH rapidly falls. pH is therefore a clear indicator of death (Aaraas et al., 2004). According to our findings, superchilled storage of oysters did not cause a decline in survival ability.

According to PCA analysis (Figure 3), the two principal factors described 94.43% of the variation between variables. The first factor, representing 73.27% of the total variation, described primarily the effect of superchilled storage on microbiological count. Accordingly, 68.4% of the variation was observed on raw oysters, and during the superchilled storage 76.5% on day 0, 10.2% on day 5, and 12.8% on day 9. In raw samples, high fecal coliforms, E. coli, and Salmonella spp. levels were related to pH. On day zero the microbial loads of fecal coliforms, E. coli, Salmonella spp. and V. cholerae non-O1/non-139 decreased significantly in non-ozonated- superchilled samples, indicating the effect of storage. However, V. cholerae levels decreased in ozonated- superchilled samples, indicating the effect of both treatments. Pearson correlation suggested a strong positive relationship between the microbiological levels of fecal coliforms and V. cholerae non-O1/non-O139 in non-ozonated- superchilled samples and pH (r = 0.976, p = 0.05; r = 0.910, p = 0.05, respectively). Thus, to improve the microbial safety of oysters it is essential to understand the physiology of bivalves during post-harvest storage, as oysters are traded as live animals.

The latest market reports in 2016 reflect an increased trade in oysters, totaling about 70,000 tons, with an estimated price of US$ 20.6 by dozen, and $US 3.3 each (Food and Agriculture Organization [FAO], 2017). According to Chen et al. (2017) an oyster industry may be an economically viable pursuit if a minimum selling price of US$1.35 per oyster can be achieved. In México, the current market price for locally grown non-depurated oysters is US$0.10 per oyster (Comisión Nacional de Acuacultura y Pesca [CONAPESCA], 2017). Considering the depuration process costs plus the cost of oysters (without the superchilling cost), we estimated that the selling price per shellstock oyster would be approximately US$0.20 at the depuration facility, US$0.44 in the central seafood wholesale market, US$0.73 in local restaurants, and up to US$1.60 in restaurants in México City, Monterrey, and Cancún. Although these prices will increase when superchilling process costs are considered, superchilled depurated oyster production would still be profitable as in recent years exports of chilled and frozen bivalves have grown worldwide (Food and Agriculture Organization [FAO], 2016).

The results of this study demonstrate that E. coli is eliminated more rapidly than V. cholerae and Salmonella spp. from the gut tissue of the tropical oyster Crassostrea virginica and thus, is an inadequate microbiological quality index as it is unable to demonstrate general depuration capacity. Ozone depuration for 6 h at 0.4 mg/L enables the efficient reduction of fecal coliforms, E. coli, Salmonella spp. and V. cholerae levels in oysters, below the maximum tolerable limits set by Mexican regulations. Superchilled storage at -1°C improved disinfection efficiency and microbiological quality of ozonated oysters at 0.4 mg/L for up to 9 days. During the storage period the 0.4 mg/L treatment attained lower fecal coliform and E. coli levels, reduced 99.5% of V. cholerae non-O1/non-O139 counts, and Salmonella spp. was not detected. Future studies are still required to assess the efficacy of this process in reducing pathogenic V. parahaemolyticus and V. vulnificus naturally accumulated in oysters. The absence of ozone depuration resulted in higher levels of fecal coliforms, E. coli, and V. cholerae non-O1/non-O139 in oyster samples during superchilled storage than those observed in control, 0.4 and 0.6 mg/L ozonated oyster samples. Superchilled storage only decreased Salmonella spp. levels in non-ozonated oysters while this bacterium was not detected in ozonated oyster samples during superchilled storage.

Combining the direct-ozone depuration pretreatment and superchilled storage has a synergistic effect and several processing advantages. The depuration time is shorter, the oyster survival is not adversely affected, and the microbiological safety of oysters improves with an acceptable safe level for human consumption up to 9 days. The use of superchilled storage at -1°C of fresh ozonated shellstock oysters is a promising technology but needs further optimization since V. cholerae, E. coli and Salmonella spp. in cold-stored oysters could potentially survive and remain as a significant health risk for raw oyster consumers. This process combination appears to be a potential technology, holding both practical and economic interest for marketing strategies, and improving the profits of local producers. It may thus represent a viable option for the oyster industry to preserve raw oysters, ensuring microbial safety throughout the supply chain to protect public health.

KL and VP contributed conception and design of the study and wrote the manuscript. SR contributed conception and design of the study. VS, IR, and RU performed the analytical techniques. VP, AF, and DM performed the statistical analysis. All authors contributed to manuscript revision, read and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer AA and handling Editor declared their shared affiliation.