In North Africa, traditional technologies for food preservation rely almost exclusively on natural hurdles (fermentation, dehydration, high osmotic pressure and/or, heating) to inhibit the growth of undesirable bacteria and provide safe and stable foods. The contribution of the resulting products to food security in North African countries is undeniably due to their availability at affordable prices, high nutritive value, and health-promoting properties (Anukam and Reid 2009; Peres and others 2012). However, their hygienic quality is highly variable depending on many factors including the microbiological quality of raw materials and ingredients, as well as the sanitary conditions during harvest, manufacture, packaging, and storage. They may thus compromise the consumer's health. Indeed, contamination of such foods with pathogens and/or microbial toxins is well documented. Table 6 provides examples of bacterial pathogens associated with the main traditional dairy, mat, and plant products. Moreover, traditional foods have been reported to be associated with infectious diseases and intoxications (Cosivi and others 1998, Belomaria and others 2007, Bendahou and others 2008) or with other pathogens of concern to food safety (Paramithiotis and others 2012). Yet, only few traditional food-related outbreaks have been recorded in North African countries due to the poor public awareness and the lack of media reports, especially when the issues of intoxication episodes are not severe or do not require emergency hospitalizations.
Traditional dairy products of North African countries are generally obtained from raw milk, with few exceptions where the milk is heated or boiled, which undergoes spontaneous acidification with LAB (Table 3) before proceeding to other technological steps depending on the desired final product. Therefore, these dairy technologies rely essentially on the fermentation with LAB as a barrier to prevent the growth of pathogenic and spoilage microorganisms. LAB bacteria have the generally recognized as safe status (GRAS) due to their long history of safe consumption by humans, and to the fact that they have rarely been associated with food intoxications or infectious diseases (Hammes and Tichaczek 1994). In addition, virtually, all species of LAB have been shown to produce a variety of biologically active substances, including organic acids, hydrogen peroxide, carbon dioxide, diacetyl, reuterins, and bacteriocins, all with antagonistic activities against microorganisms of health and spoilage significance (Piard 1992; Ouwehand 1998). Moreover, proteolytic LAB have been shown to generate endogenous bioactive peptides with potent antimicrobial activities upon hydrolysis of milk proteins during fermentation (Hayes and others 2006; Benkerroum 2010; Ghalfi and others 2009). Nonetheless, studies on traditional North African dairy products have revealed that those among them which rely only on the lactic acid fermentation and the consequent lowering of the pH (3.8 to 4.5), such as fermented milk products including zabadi, raib, lben, and fresh cheeses are generally of poor microbiological quality, as suggested by the high counts of indicator bacteria, fecal coliforms, and enterococci, which exceed 104 colony forming unit (CFU)/mL or g (Hamama 2002), and the occurrence of serious foodborne pathogens (Table 6). In addition, these products have a short shelf life (3 to 10 d) even when stored at refrigeration temperature (Samet-Bali and others 2009), suggesting that they are prone to support the growth of various undesirable microorganisms. To circumvent such a limitation and enhance the safety and keeping quality of North African traditional dairy products, other hurdles to microbial growth including salting and/or drying (low aw) or heating have often been combined with lactic acid fermentation. Examples of such products are the Moroccan-brined cheese jben malah (matured and stored in saturated brine), Egyptian domiati and tallaga cheeses (5% to 10% salt content and stored in brine), dried kishk, and Algerian aoules obtained from spontaneous acidification of milk followed by heat treatment and sun-drying. The pH of these products is as low as 3.5 and their dry matter increases to levels as high as 90% to 92%, corresponding to water activity (aw) values of 0.34 to 0.43 where microorganisms can no longer survive (Steinkraus 1995; FAO 1990; Tamime and McNulty 1999; Mennane and others 1973). Addition of aromatic plants (thyme, oregano, and rosemary) along with high salt concentration to some traditional dairy products, such as the Moroccan smen, has also been practiced empirically to add a safety factor and to avoid surface spoilage with molds, as these plants are well known for their potent antifungal activities (Cowan 1999; Hammer and others 1999; Chebli and others 2003; Rota and others 2004; Amarti and others 2008). In spite of these treatments, pathogenic bacteria and molds have been detected in virtually all North African traditional dairy products, and some of these bacteria would grow and produce highly toxic metabolites under certain conditions. Survival of raw milk pathogens to the processing steps or contaminations due to the lack of good manufacture practices (GMPs) and personal hygiene, or improper storage conditions, are the most frequent causes for the presence of undesirable microorganisms in finished products. Moreover, the poor sanitary conditions during milking and the lack of veterinary care for traditional small herders (the prevailing management system of husbandry in North Africa), in addition to the low hygienic quality of water, contribute to produce raw milk of poor microbiological quality. Regardless of the type of domestic milk animal (cow, sheep, goat, buffalo, and camel), the total viable counts usually exceed 106 CFU/mL and the counts of fecal indicators (coliform and enterococci) are higher than 104 CFU/mL (Hamama 2002; Benkerroum and others 2003a; El-Diasty and El-Kaseh 2007). Moreover, foodborne pathogens of major concern in food safety such as Listeria monocytogenes, Mycobacterium bovis, Mycobacterium tuberculosis, enterohemorrhagic Escherichia coli, Staphylococcus aureus, and Campylobacter jejuni have frequently been isolated from raw milk in North African countries (Hamama 2002; El Marrakchi and others 1993; WHO 1994; Cosivi and others 1998; Benkerroum and others 2004b; Bendahou and others 2008). Therefore, the frequent presence of pathogens in fermented milks or cheeses obtained from such raw milk is not surprising. When raw milk is heavily contaminated and the technological processes are carried out under poor sanitary conditions, as is usually the case in North African dairy farms, natural hurdles to microbial growth are a limited safety factor (Benkerroum and Tamime 2004).
Another issue of concern related to North African traditional dairy products is the presence of enterococci at high levels, usually exceeding 104 CFU/mL, with Enterococcus faecalis and foecium as the predominating species (Benkerroum and others 1984; Tantaoui-Elaraki and El Marrakchi 1987; Benkerroum and Tamime 2004). The presence of this group of microorganisms in fermented foods is, in fact, controversial. Although members of Enterococcus genus have been shown to possess highly desirable technological and health promoting properties (Bertolami and Farnworth 2003; Belamri and Benkerroum 2005; Pollmann and others 2005; Moreno and others 2007; Zeyner and Boldt 2006), their association with food spoilage (Franz and others 1996), food intoxication (Giraffa and others 2006; Gardini and others 2007), nosocomial infections (Kayser 2003), and spreading of antibiotic resistance through the food chain (Valenzuela and others 2008) has also been well documented. In a survey on risk factors in enterococcal strains isolated from Moroccan dairy products, Valenzuela and others (2008) revealed the widespread multiantibiotic resistance and/or occurrence of virulence factors (sex pheromones, collagen adhesins, enterococcal endocarditis antigen, and enterococcal surface proteins) among the isolates of Ent. faecalis and Ent. faecium (Valenzuela and others 2008). According to these authors, high counts of enterococci represent a risk factor in Moroccan foods, and appropriate measures should be taken to reduce their incidence.
In addition to food safety concerns related to bacteria and fungi, the presence of viruses, parasites, and protozoa in North African traditional dairy products may also represent serious hazards to consumers, which has so far been disregarded or overlooked. The occurrence of the latter microorganisms in North African raw milk has been repeatedly demonstrated (Carter 2005; Dawson 2005; Pollmann and others 2005) and is of paramount importance to the safety of dairy products, especially when no heat treatment is applied during processing or at point of consumption. In view of the high incidence of zoonotic diseases in North African herds (Araba and Essalhi 2007; Berrag and others 2009; Bouzid and others 2010) especially with the lack of veterinary care in small farms and the poor hygienic quality of water (the probable vehicle for various viruses, protozoa, and parasites), this issue warrants due attention from all stakeholders and scientists.
The presence of mycotoxins in traditional North African dairy products also raises an increasingly alarming concern regarding public health safety. It is well established that mycotoxins may contaminate dairy products by molds growing on them under certain conditions, or by the carryover of mycotoxins occurring in animal feedstuffs ingested by dairy animals and later transferred from blood into milk (van Egmond 1983; Veldman and others 1992; Galvano and others 2001; Zinedine and others 2006; Masoero and others 2006). The most important of such mycotoxins to dairy products is aflatoxin M1 (AM1) derived from the conversion of ingested aflatoxin G1 in the animal liver (Tantaoui-Elaraki and Khabbazi 1984). AM1 was detected at concentrations ranging between 4.0 and 6.0 μg/L in 12.5% of Libyan raw milk (El-Diasty and El-Kaseh 2007). In a study on the occurrence of AM1 in raw milk marketed in different cities of Libya, Elgerbi and others (2004) showed that 35 samples of raw bovine milk out of a total of 49 (71.7%) tested positive for AM1, with average concentrations varying between 0.12 and 0.72 μg/L depending on the city from which the samples were taken; 34 among the 35 positive samples (about 97%) were contaminated with levels exceeding the maximum tolerable level (MLT) of 0.05 μg/L (Elgerbi and others 2004). Higher concentrations (average of 6.3 μg/L) were reported in Egyptian raw bovine milk (El-Sayed and others 2000). Such levels of contamination with AM1 are alarming in view of the MLT which should be less than 0.05 μg/kg or L, depending on the country and commodity (Dohlman 2003). In Moroccan pasteurized milk, AM1 was detected at high frequencies (88.8%), but at relatively low concentrations compared to those found in Libyan and Egyptian milks; the concentrations reported ranged between 0.001 and 0.117, with an average of 0.018 μg/L, and 7.4% of the samples contained higher levels than the MTL (Zinedine and others 2007a; Zinedine and Mañes 2009). According to Tantaoui-Elaraki and Khabbazi (1984), AM1 concentration in raw milk increased by 3.2- to 3.7-fold in cheeses made from contaminated milk, as neither heat treatment nor the subsequent steps in cheese making (curd drainage, handling, and maturation) removed much of the toxin initially present in the milk. It could be anticipated, on this basis, that levels of AM1 in cheeses obtained from contaminated milk would exceed the MTL even when the initial level of AM1 in raw milk is below this value. However, a survey on the occurrence of AM1, aflatoxin B1 (AB1), and aflatoxin G1 (AG1) in Egyptian dairy products revealed that AM1 was detected in raw milk more frequently and in more elevated amounts than in different cheese varieties or in dried milk; AM1 concentrations of 6.3, 5.0, 6.0, and 0.5 μg/L or kg were detected in raw milk, dried milk, hard cheese, and soft cheese, respectively, whereas AB1 and AG1 were detected in hard cheese at concentrations of 3 and 6 μg/kg, respectively (El-Sayed and others 2004). Similarly, the average concentrations of AM1 in Libyan white soft cheeses (from 0.21 to 0.34 μg/kg) were lower than those determined in raw milk; yet, 15 cheese samples out of 20 (75%) tested positive for AM1 and contained concentrations exceeding the MTL (Elgerbi and others 2004). It could be argued, however, that there is no correlation between the levels of mycotoxins in the cheeses samples analyzed in the latter studies and those determined in the raw milk, as all the samples were taken at random from local markets. In a more relevant study, Hassanin (2006) has monitored the level of AM1 in yogurt, yogurt-cheese, and acidified milk produced from a naturally contaminated milk with AM1. The results revealed that the concentration of AM1 significantly decreases as a function of time during storage at refrigeration temperature, which was explained by the interference of LAB with mycotoxin activity. Many studies have shown that LAB are able to sequester, or inhibit, the in situ production or toxicity of mycotoxins, thereby reducing their potential risk in fermented milks and cheeses (Gourama and Bullerman 1999; Kim 1988; Bianchini and Bullerman 2009). Dairy lactic acid bacteria belonging to different genera have been reported to be effective in removing AB1 and AM1 (El-Nezami and others 1998; Oatley and others 2000; Pierides and others 2000). Nonetheless, further studies are needed to accurately estimate the concentrations of mycotoxins and monitor their fate during processing and/or storage in different traditional North African dairy products obtained from the milk of different animal species. Such studies are crucial to characterize the risk and estimate the dose/response in a risk-assessment process, the cornerstone in any future regulatory or control measure related to food safety worldwide.
The overall safety of meat products is contingent on many factors including the initial quality of meat and ingredients, the sanitary conditions during handling, processing, and storage, as well as the preservation hurdles used. Raw and offal meat used to produce North African traditional meat products is generally heavily contaminated with microorganisms of hygienic significance. In this regard, a study conducted by Cohen and others (2006) on beef, lamb, and beef-offal marketed in the city of Casablanca (Morocco) has shown that these products were highly contaminated; and pathogenic staphylococci and Cl. perfringens were detected at relatively high counts (Table 7). A similar study on ground beef marketed in the city of Fez (Morocco), carried out by Oumokhtar and others (2008) also demonstrated the poor sanitary quality of the product (Tables 7), with more than 80% of the analyzed samples not meeting the Moroccan regulatory standards for meat products (http://www.onssa.gov.ma/onssa/fr/vv_dec15.php). In the latter study, Salmonella sp. and Shigella were detected in a 25 g sample at the respective frequencies of 17.5% and 2.5%. Furthermore, pathogens such as enterohemorrhagic E. coli, Yersinia enterocolitica, L. monocytogenes, and Salmonella spp. have been repeatedly isolated from meat samples in Morocco and other North African countries (Karib and others 2003; Ettriqui and others 1995; Abdul-Raouf and others 1996; Al-Gallas and others 2002; Khosrof Ben Jaafar and others 2002; Benkerroum and others 2004b). Occurrence of parasites (protozoa and helminthes) of the genera Toxoplasma, Sarcocystis, Cryptosporidium, Fasciola, Flatworms, Tapeworm, and roudwarms in meat and offal is also well documented (Sawadogo and others 2005; Valinezhad and others 2008; Abdel-Ghaffar and others 2009; Berrag and others 2009). Given the poor hygienic quality of raw meat, traditional technologies in developing countries use more than one hurdle, acting in synergy, to ensure a relatively satisfactory degree of hygienic quality. Indeed, almost all the traditional technologies for meat transformation in North African countries combine salting, herb and spice addition, drying, and, occasionally, cooking, especially when long shelf life and a high degree of safety are sought. Less stringent conditions are used when a product is not intended for an extended preservation period and is cooked before consumption, as is the case for merguez and tehal. In cured and fermented meat products, such as sujuk, gueddid, kourdass, and pastirma, salting, drying, and herb and spice adjunction are the main parameters used to ensure their safety and stability. While salting and/or drying reduce the water activity to levels below 0.86 where no pathogenic bacteria would grow (Jay 2008; Barbosa-Cánovas and others 2003), spices inhibit specific microorganisms including bacteria, molds, protozoa, and viruses (Farag and others 1989; El-Khateib 1997; Cowan 1999; Daferera and others 2000). In addition, a decrease in the pH to about 5.5 at the first phase of maturation, while not efficient by itself to inhibit many pathogens, stimulates the growth of LAB, which, in turn, will further inhibit undesirable microorganisms through antibiosis interactions. In this regard, the wide occurrence of bacteriocin-producing enterococci has been reported in Tunisian gueddid (Ben Belgacem and others 2008), and the protective effect of bacteriocins in meat systems has been demonstrated (Benkerroum and others 2008, 2005). In addition to these conditions, some traditional North African meat products, such as the Moroccan khlii and Libyan ban-shems, the cooking during processing inactivates microbial pathogens or toxins that may be present on the meat or offal used for their preparation.
Table 7. –Average microbial counts (log CFU/g) in different meat and offal samples in 2 major Moroccan cities
Although there is a lack of scientific data regarding the hygienic quality of North African traditional meat products and epidemiological studies on their involvement in food outbreaks, it could be anticipated that they may compromise food safety as suggested by the widespread occurrence of pathogens in North African traditional meat products (Table 6), and also for the main reasons below:
- Lack or inappropriate veterinary care in the farms where meat-production animals are raised; a weak prophylactic program and inadequate treatment of diagnosed bacterial infections or parasitic infestations (Berrag and others 2009).
- Slaughtering, carcass dressing, evisceration, and meat cutting are generally done in poor sanitary conditions which, combined with the nonrigorous or absent (farm-slaughtering) veterinary inspection at slaughter, strongly suggest that meat or offal deriving from animals infected with virulent bacteria, viruses, or parasites would be used to manufacture traditional meat products.
- Production of traditional meat products in small farms, butcheries, or food shops where the sanitary conditions are usually not appropriate, and none of the GMP, good hygiene practice (GHP), or hazard analysis and critical control point (HACCP) program is implemented.
- Inadequate conditioning, packaging, and storage conditions. Yet, efforts in packaging are being increasingly made as part of a marketing approach.
In fact, it is well established that these are the main factors that impact the safety and keeping quality of the finished meat products, and failure to address any of them properly will invariably result in a meat product of high risk to consumers. Nonetheless, on the basis of moisture content, North African traditional meat products may be divided into 3 groups as defined by Marshall and Bal'a (2007) and Leistner (2008), each of which present a different pattern of health risks to consumers from the microbiological standpoint: (i) dry meat products (less than 15% moisture) such as gueddid, kurdass, khlii, and ban-shems, (ii) intermediate-moisture meat products (15% to 20% moisture or an aw of 0.65 to 0.90) typically represented by pastirma and certain types of sujuk (Table 4), and (iii) fresh meat products (>20% moisture) including merguez, mkila, tehal, and some types of sujuk where no or partial drying is applied during processing. Due to their low moisture and/or high salt contents, the first 2 groups are generally regarded as microbiologically safe, and they may be consumed as such or after being lightly cooked; these include khlii, some sujuk types, and pastirma. Such an assumption has been substantiated by some studies, showing that the overall microbiological counts are either low or dominated by beneficial LAB and that some foodborne pathogens do not grow or survive in these products (El-Khateib 1997; Bennani and others 2000; Huang and Nip 2009; Kalalou and others 1999). El-Khateib (1997) showed that the numbers of Enterobacteriaceae, and yeasts and molds were less than 100 CFU/g in 50 samples of Egyptian pastirma, which were also free from Salmonella. The absence of Salmonella was explained by the inhibitory effect of the spice paste used to cover pastirma, as has been demonstrated in vitro (El-Khateib 1997). The same study showed that the total aerobic count (TAC) and the Lactobacillaceae ranged between 1×104 and 9×106 CFU/g, suggesting that LAB are the main responsible for the evolution of the product during ripening, which represents a good indication regarding the safety of the product. In addition to the antimicrobial effect of the cover paste, inhibition of pathogenic bacteria in pastirma was suggested to be due to the combined effect of water loss and salt content (4.5% to 6%) with the consequent decrease in water activity (Leistner 2000b; Bechtel 2001). In effect, challenge studies between an E. coli O157 : H7 strain and protective cultures of Lb. sakei and Staph. xylosus in pastirma have demonstrated that the most significant reduction in the counts of the pathogen was recorded after the drying step regardless of whether or not the protective cultures were present (Aksu and others 2008). Similar results were reported for Moroccan gueddid where numbers of TAC, coliforms, and staphylococci showed a dramatic decrease after the maturation step to reach an undetectable level in a 1 g sample for coliforms and staphylococci, and about 40 CFU/g for TAC (Kalalou and others 1999). According to their study, the sharp decrease in the microbial counts paralleled the decrease in water activity (aw) to a final value of 0.66. Furthermore, neither Salmonella nor clostridia were detected in laboratory-made or commercial gueddid samples (Bennani and others 1995; Bennani and others 2000). However, despite such reassuring data, though partial and fragmentary, there is no absolute guarantee for the safety of these products. Indeed, the related salted-dried jerky meat prepared in a similar manner and having a water activity value as low as 0.3 has repeatedly been associated with a number of Salmonella and Staph. aureus outbreaks in the U.S.A. (CDC 1995; Eidson and others 2000; Smelser 2004; Allen and others 2007). Nonetheless, it could be assumed that the marinated variant of gueddid would be microbiologically safer than the classical type. The application of acid marinade to meat before drying has been shown to enhance significantly the microbiological quality of the final product (Nummer and others 2004). On the other hand, North African traditional fresh meat products are usually heavily contaminated with microorganisms of health and spoilage significance; they thus present a higher risk to consumers than their dry or intermediate-moisture counterparts. However, such risk may be reduced at consumption, as these products are cooked or grilled before consumption. In this regard, a study on the microbiology of different Egyptian fresh sausages showed that the aerobic plate count (APC) and Enterobacteriaceae counts ranged from 1.1×104 to 1×108 and from 1×102 to 1×107 CFU/g, respectively, and Cl. perfringens and coagulase-positive Staph. aureus were detected at the respective frequencies of 26% and 29% (El-Khateib 1997). Moreover, in a study on the incidence of shiga toxin-producing E. coli O157 in Moroccan meat products, the pathogen was detected in 20% of the spiced ground meat normally used in merguez preparation, but not in merguez sausages; such a discrepancy could be explained by the sampling procedure and the limited number of samples studied (Benkerroum and others 2004b).
Mycotoxins are contaminants of microbial origin, which also raise concern about the safety of meat products. The presence of molds in meat and meat products is well documented, and under certain conditions, they may grow and produce mycotoxins (Sweeney and Dobson 1998). Molds usually grow on dry or intermediate-moisture meat products during the first days of drying, especially if the drying process is slow or the relative humidity in the atmosphere is high. They may also grow in the finished product if the storage conditions are not adequate. However, this growth is usually considered by the producers only as a harmless surface discoloration and is then removed by brushing the sausages or cleaning them with a wet cloth to avoid wasting meat. Yet, such growth may represent a serious risk factor if the contaminating molds are toxinogenic. A study on the mycology of pastirma has revealed the predominance of species belonging to Penicillium and Aspergillus genera (Abdel-Rahman and others 1984). These genera are well known for their ability to produce mycotoxins (Sweeney and Dobson 1998). A study on Egyptian pastirma showed that the numbers of molds varied from 103 to 106 CFU/g in summer and from 102 to 105 CFU/g in winter, and that Aspergillus, Penicillium, Mucor, Rhizopus, Fusarium, and Cladosporium were the predominating genera in the product (Refai and others 2004). In addition, the spices used in the preparation of North African traditional meat products are usually contaminated with variable levels of mycotoxins and are thus potential sources for the contamination of these products. Aflatoxins were determined in the pastirma spice paste and its constitutive spices individually; the contamination level of the spice paste varied from 9.6 to 120 μg/kg, and in pepper (285.6 μg/kg), garlic (224.4 μg/kg), fenugreek (194.2 μg/kg), coriander (166.4 μg/kg), and capsicum (42.4 μg/kg) (Refai and others 2004). These concentrations largely exceed the maximum tolerable limit of aflatoxin B1 in spices (5.0 μg/kg) according to the European regulations (Zinedine and Mañes 2009). Therefore, it might be anticipated that pastirma and related North African traditional meat products represent a serious health risk factor associated with the consumption of these products. Pastirma was, indeed, reported to contain aflatoxins at levels varying from 2.8 to 47 μg/kg (Refai and others 2004).
Fruits and vegetables are contaminated by a wide variety of microorganisms including bacteria, fungi, viruses, and protozoa of different origins. These contaminations may occur in the field (soil, manure, compost, wastewater sludge, irrigation water, equipment, workers, animals, and so on), during postharvest operations (conditioning, packaging, and distribution), or at the household prior to consumption (Burnett and Beuchat 2001; Matthews 1983). Therefore, microorganisms initially present in fruits and vegetables are highly variable in numbers and in nature, and are generally predominated by saprophytic bacteria and molds, considered as the resident microflora, that do not raise serious health concerns (Jay 2008; Badosa and others 2000). Nonetheless, fruits and vegetables have extensively been shown to be contaminated with a variety of pathogenic bacteria, protozoa, and viruses (Beuchat 1998; Badosa and others 2000; Robertson and Gjerde 2000; Buck and others 2003; Bhagwat 2006; Matthews 1983), which is regarded as a risk factor for public health. Indeed, the increase in consumption of fruit and vegetable that has been recorded worldwide in the last 2 decades has been paralleled by an increase in foodborne disease outbreaks attributed to fresh produce (Beuchat 1996; Tauxe and others 1997; Buck and others 2003). This was corroborated by the association of a number of fruits and vegetables from around the world with gastroenteritis outbreaks (Buck and others 2003; Matthews 1983), most of which were of bacterial origin, with Salmonella and E. coli O157 : H7 as the primary etiological agents (Heaton and Jones 2008). On the other hand, some microbial groups among the resident microflora play a key role in the transformation of vegetables into more stable, safe, and palatable products when fresh produce is preserved for consumption out of season. Among the beneficial of these epiphytic microorganisms in fruits and vegetables, LAB, and to a lesser extent yeasts, are responsible for spontaneous fermentation of a number of vegetable products such as olives and various vegetables. Natural fermentation has long been used on an empirical basis in the preservation of vegetable products, and in particular, in the case of olives, it is even necessary to make the fruit edible. In fact, natural olive fermentation fulfills 2 main objectives: (i) inhibition of microorganisms of health and spoilage significance, thereby reducing health risks and product losses, and (ii) making olives palatable, essentially by removal of the bitter glycoside euloropein, production of various aroma compounds, and softening, to some degree, the flesh of the fruit during fermentation.
As mentioned earlier, vegetable products are traditionally preserved in the North African region by lactic acid fermentation combined with salting (dry salt or brine), direct acidification (addition of vinegar) or, in few instances, by sun-drying. Conversely, fruits such as grapes, figs, and prickly pears are essentially dried, as the loss of moisture combined with the consequent increase in sugar concentration in the fruit results in a sharp decrease in water activity, thereby restricting the growth of spoilage and pathogenic microorganisms. The high diversity of North African traditional vegetable products and the raw material from which they derive, in addition to the variability in the technological processes and the sanitary conditions used for their manufacture, make the microbiological characteristics of the finished products and related risk factors also variable. Nonetheless, the fruits and vegetables eaten raw or after being transformed by traditional technologies are regarded as safe and wholesome for most North African consumers. Yet, this presumed safety has not been substantiated by microbiological and epidemiological studies, a situation that reflects the lack of awareness of the health risks associated with the consumption of vegetable products in these countries. Furthermore, because of the lack of foodborne disease investigations and surveillance, most disease outbreaks related to the consumption of vegetable products remain undetected and insufficiently reported in the scientific literature. Also, the level of contamination and the incidence of pathogenic microorganisms of/in fruits and vegetables are anticipated to be higher in North African countries than in developed countries. The lack of implementation of quality assurance programs including GAP, GMP, and HACCP throughout the entire production and distribution chain (pre- and postharvest, transportation, handling, and so on) increases the health risk associated with the consumption of such products. The risk is even greater as untreated urban wastewater and sewage sludge or manure continue to be used for irrigation or fertilization (Bazza 2003; Ghazy and others 2004). This practice, though recognized to be illegal, is tolerated by regulatory authorities in North African countries, and is widely used. In Morocco, about 70 million m3 of untreated urban wastewater are used annually to irrigate some 7000 ha of fruit orchards and vegetables as an alternative fresh water source, and as an easy and economical way for disposing wastewater (Bazza 2003; Choukr-Allah 2004). Similarly, the capacity of wastewater-treatment facilities in Egypt is either short or produces insufficiently treated wastewater (Ghazy and others 2004); therefore, large amounts of wastewater and sewage sludge are used to irrigate and fertilize fruit orchards and vegetables in the country. The use of raw wastewater in agricultural activities has been demonstrated to increase the potential of the resulting crops to spread bacterial or parasitic diseases (Ait Melloul and Hassani 1999; Ait Melloul and others 2002). A bacteriological analysis of various vegetables obtained from several wastewater-irrigated agricultural regions in Morocco showed high counts of TAC (>9 log10 CFU/g) and fecal indicators (total-coliforms, fecal-coliforms, and enterococci)(>5 log10 CFU/g), suggesting that the consumption of these vegetables put consumers at high risk (Ibenyassine and others 2009). The same study showed that opportunistic Gram-negative pathogens of the Enterobacteriaceae family (Citrobacter freundii, Enterobacter cloacae, E. coli, Klebsiella pneumoniae, and Serratia liquefaciens) were detected in the studied vegetables at frequencies varying from 11% to 28%, with Enterobacter sakazakii (12%) and Salmonella arizona (0.7%) being the pathogens of the most concern to the safety of these crops. Moreover, in a Moroccan region (Al Haouzia, Marrakech) where untreated wastewater spreading is widely practiced for the production of vegetables such as lettuce, tomatoes, parsley, and potatoes, as well as cereals, the prevalence of Salmonella infection (32.56%) in the community living in the region and consuming locally produced crops has been shown to be significantly higher than that recorded in a control region where no wastewater spreading is practiced (1.14%). Another study carried out in the same region revealed that the prevalence of protozoan infections (giardiasis and amebiasis) among children of the wastewater-irrigated region was significantly higher (72%) than that recorded in the control area (45%) (Ait Melloul and others 2002).
Traditional food products of plant origin in North African countries are generally obtained from locally produced crops with inconsistent sanitary quality. Therefore, the safety of finished products is largely dependent on the efficacy of natural hurdles to inhibit or inactivate undesirable microorganisms initially present in the raw material. Olives are the most important vegetable products that are transformed by traditional technologies in North African countries; they are either brined (green olives) or dry-salted (black ripe olives). In both cases, they undergo spontaneous lactic acid fermentation, although in the latter case, the fermentation is greatly retarded by the high salt concentrations used for their preservation (8 to 14 g salt per 100 g olives). Other vegetables such as peppers, onions, carrots, string beans, chili, and cauliflower are either brined in a similar manner as for green olives (Steinkraus 1995) or salted and “packed” into an acid solution (usually vinegar) of a pH value below 3.0 as unfermented (fresh-pack) pickles. In the fresh-pack pickles, the raw material is usually heated or soaked in hot water for few minutes to reduce the overall microbial load of the product before pickling. Therefore, it would be reasonable to expect that these unfermented pickles do not raise serious health concerns essentially due to the low pH of the brine in addition to the heat treatment where microbial growth is very unlikely to occur (Gómez and others 1988). They would therefore be categorized as low-risk vegetable foods. Conversely, in fermented vegetable products, faulty fermentation is not uncommon and pathogenic microorganisms may grow or survive in the finished products (Fleming and others 1985; Caggia and others 2004). In fermented pickled vegetable products, microbial competition, acidity, and reduced water activity (high salt content) are the main parameters that inhibit undesirable microorganisms to safeguard the health of consumers. However, the usual salt concentration used in the brine (5 to 7 g salt/100 mL) is not high enough to reduce the water activity to levels that would strongly inhibit the growth of undesirable microorganisms. In addition, when more than 8% (w/v) of salt is used in the brine, the growth of LAB is also retarded significantly and the pH remains relatively high at the end of fermentation (about 4.5) (Fernández and others 1997), conditions that provide an opportunity for salt-tolerant or halophilic pathogens and spoilage microorganisms to grow during the early stages of fermentation. Staphylococci grow well at salt concentrations between 7% and 10% and a low pH of 4.2, and other pathogens such as E. coli O157 : H7, L. monocytogenes, and Salmonella spp. have been shown to develop resistance under stressful conditions, including low pH (Gahan and others 1996; Lou and Yousef 2003; Beales 2004; Lee 2000a). Therefore, in the fermentation of green olives, the usual salt concentration used does not exceed 7 g salt/100 mL brine to allow good growth of fermentative LAB. These bacteria, which represent only a small proportion of the initial epiphytic microflora of the product, should rapidly outgrow the competing microbial groups and produce sufficient acidity to reach a pH of about 3.5 to 3.8, and then the resulting product may be considered reasonably safe.
LAB have been shown to predominate throughout the entire fermentation period (90 d) in naturally fermented Algerian green olives, although yeasts were consistently present at relatively high counts (3 to 6 log10 CFU/g of olives) as a secondary microflora (Kacem and Karam 2004). These authors have shown that the counts of LAB increased steadily since early phases of fermentation to reach about 7 log10 CFU/mL after 90 d of fermentation. A significant increase in yeast counts has also been noted, starting from day 60 of fermentation, in a typical behavior of a secondary fermentation flora (Fleming and others 1985). Meanwhile, the counts of coliforms, staphylococci, and psychroptrophs were reduced to different extents (Table 8). Such a decline may be attributed not only to the effect of pH and salt content of the Algerian fermented green olives (Kacem and Karam 2006b), but also to inhibitory substances inherently present in olives or to specific metabolites produced by the predominating species of LAB and yeasts during fermentation. Euloropein has been shown to inhibit, to different extents, various pathogens including Salmonella typhi, Salmonella Enteritidis, Vibrio parahaemolyticus, Staph. aureus, and Vibrio cholerae (Tassou and Nychas 1994; Tassou and Nychas 1995; Bisignano and others 1999). Among LAB isolated from fermented green olives in North African countries (Table 9) and elsewhere (Fleming and others 1985), Lb. plantarum has been consistently reported to be the predominating species throughout the fermentation period (Kacem and others 2005; Chamkha and others 2008); and strains of this species have extensively been shown to produce bacteriocins active against Gram-positive and Gram-negative bacteria (Kacem and others 2005, 2003; Dobson and others 2012). Likewise, yeasts have been reported to be consistently present at elevated counts (> 6 log10 CFU/g) in North African fermented olives (Asehraou and others 1992; Asehraou and others 2000; Kacem and Karam 2006b; Hernández and others 2001), and to contribute to their safety. It is indeed well established that yeasts produce aroma compounds including organic acids, diacetyl, ethanol, and other metabolites resulting from lipolytic activities that also possess antimicrobial activities, thereby contributing to flavor development as well as to the safety improvement of foods (Hernández and others 2001; Arroyo-Lopez and others 2008). Furthermore, among the major yeast species isolated from fermented olives (Table 9), the so-called killer yeasts tend to predominate owing to their ability to produce a “killer toxin” essentially active against other competing yeasts (Llorente and others 1997; Asehraou and others 1992; Hernández and others 2001; Maqueda and others 1998), but could also inhibit various pathogenic Gram-positive bacteria (Izgü and Altinbay 1986). Therefore, killer yeasts would not only reduce the incidence of bloater defect (Asehraou and others 2000), but also contribute to the enhancement of the safety of finished products. Nonetheless, the safety of naturally fermented olives may not rely only on the above-mentioned safety factors, as the degree of protection they offer vary widely depending on the fermentation parameters and from a batch to batch (Asehraou and others 2000; Kacem and Karam 2004). In effect, foodborne pathogens of concern to food safety have been isolated or shown to survive from/on Spanish-style green olives or Greek-style black olives (Table 6), and cases of botulism and other food poisoning disease due to consumption of fermented green or black olives have been reported (Okudaira and others 1962; Pereira and others 2008; Pingeon and others 2011; Anon 2012). The competitiveness of the predominating LAB and yeasts as determined by their ability to produce inhibitory metabolites or competition for nutrients, and the initial bacteriological quality of olives and brine, in addition to the sanitary conditions during manufacture, are the main parameters that determine the safety status of traditional fermented olives. Presently, olives are heat-treated before or after preservation and various chemical food-grade additives such as acids, sorbates, and benzoates and parabens are added to pickled or dry-salted olives in North African countries in order to enhance their safety and keeping quality. In the market place, the latter products are known as “romy,” litrally meaning “coming from Rome” but commonly used to refer to modern and sophisticated products, as opposed to “baladi/beldi” products (products of the country, in Arabic). It is worth mentioning, in this regard, that the “baladi” products are the most preferred by consumers though less attractive in appearance and their hygienic quality is questionable as compared to the “romy” ones. Further studies should be carried out on the hygienic quality of different commercial and home-made fermented green olives samples, and from different geographical locations of North African countries, in order to provide a sound conclusion regarding sanitary quality of naturally fermented olives and to allow an accurate assessment of the health risk associated with their consumption. Particular attention should be given to the occurrence of Cl. botuminum and its toxin, since this is considered as one of the most important safety issues in fermented olives worldwide, although the occurrence of Cl. botulinum in fermented olives appears to be rare, and only very few outbreak intoxications due to botulism of type B have been recorded (Pereira and others 2008; Anon 2012). In addition, the botulism toxin formation is unlikely at pH and aw < 4.8 and 0.94, respectively (Odlaug and Pflug 1979; Briozzo and others 1986); values of these parameters are generally lower in naturally fermented olives.
Table 8. Monitoring the counts (logCFU/g) of microbial groups in Algerian green olives during fermentation. After Kacem and Karam (2006b)
| ||Fermentation period (days)|
|Total aerobic count||4.52||7.76||6.88|
|Lactic acid bacteria||3.8||6.55||6.96|
Table 9. Predominating species of LAB and yeasts in North African naturally fermented green olives. Data are adapted from the references to present microbial species representing more than 70% of the total identified isolates
| ||Microbial group|| |
|Country of origin||LAB||Yeasts||Reference|
|Algeria||Lb. casei||Sac. cerevisiae||Kacem and Karam 2006b|
| ||Lb. paracasei||Candida parapsilosis|| |
| ||Lb. plantarum|| || |
| ||Ent. faecium|| || |
| ||Lc. lactis||NA||Kacem and others 2004|
| ||Lb. plantarum|| || |
|Morocco||NA||Sac. cerevisiae, Pichia anomala||Asehraou and others 2000|
|Tunisia||Lb. plantarum||Pichia membranaefaciens||Chamkha and others 2008|
| ||Lb. collinoides|| || |
In addition to the bacteriological hazards discussed above, mycotoxins represent a major issue regarding the safety of traditional vegetable products in North Africa. Among these products, table olives and dry fruits such as figs and raisins have repeatedly been reported to be contaminated with various mycotoxins (Table 10). The high salt contents (5 to 12 g salt/100 g product) in table olives select for the halophilic or salt-tolerant molds especially, those of the genera Penicillium and Aspergillus. Similarly, raisins and dried figs have been shown to be frequently contaminated with mycotoxin-producing strains of A. flavus and A. niger due to the osmophilic character of these species (Rao and Kalyanasundaram 1983; Pitt and others 2009; Selouane and others 2009).
Table 10. Incidence of mycotoxins (μg/kg) in selected vegetable products in some North African countries
|Commodity||Mycotoxin||Frequency (%)||Range (Average)||Origin||Reference|
|Table olives||OTA||36||0.62 to 4.8 (1.43)||Morocco||Zinedine and others 2004|
| ||OTA||100||up to 1.02||Morocco||El Adlouni and others 2006|
| ||AFB||100||up to > 0.5||Morocco|| |
| ||CIT||80||0.45 to 0.52 (0.5)||Morocco|| |
| ||OTA||–||46830*||Tunisia||Maaroufi and others 1995a|
|Figs||AFB1||5*||0.28||Morocco||Juan and others 2006a|
| ||AFG1||30||0.28 to 32.9 (8.70)||Morocco|| |
| ||OTA||65||0.03 to 1.42 (0.33)||Morocco||Zinedine and others 2007b|
| ||OTA||100||60 to 120||Egypt||Zohri and Khayria 2007|
|Raisins||AFB1||20||3.2 to 13.9 (10.7)||Morocco||Juan and others 2008|
| ||OTA||1||250*||Egypt||Youssef and others 2000|
| ||OTA||35||0.05 to 4.95 (0.96)||Morocco||Zinedine and others 2007b|
| ||AFB1||2||300||Egypt||Youssef and others 2000|
| ||AFT, CIT, OT, PAT, STG, DAX, T-2 toxin, ZEA||0||ND||Egypt||Zohri and Khayria 2007|
Contamination of table olives with various mycotoxins is well documented, and Greek-style black olives are the most incriminated (Maaroufi and others 1995b; El Adlouni and others 2006). The high salt content (about 7% to 12% g salt/g olive) inhibits the growth of almost all competing bacteria (Asehraou and others 1992; Efstathios 2006), thereby providing an opportunity for the salt-tolerant molds to grow and possibly produce mycotoxins. Among these, the most frequently encountered in North African table olives are species of the genera Aspergillus and Penicillium (Tantaoui-Elaraki and Le Tutour 1985; Gourama and Bullerman 1995; Maouni and others 2012), and the most frequently detected mycotoxins are OTA, citrinin (CIT), and AFB (El Adlouni and others 2006; Zinedine and Mañes 2009). Nonetheless, studies have shown that black olives are not good substrates for mycotoxin production (Gourama and Bullerman 1995; Eltem 1996). In this regard, Tantaoui-Elaraki and Le Tutour (1985) have demonstrated the inability of A. flavus and A. ochraceus to produce detectable amounts of mycotoxins in Moroccan table olives, while the same strains had been shown to produce high concentrations of mycotoxins in laboratory media. Similar observation has been made by Leontopoulos and others (1997) who have demonstrated that a toxinogenic strain of A. parasiticus was unable to produce AFB1 in damaged black olives originating from Greece. Similarly, Gourama and Bullerman (1995) showed that A. flavus did not produce AFB1 in pastes made from Moroccan Greek-style black olives. Furthermore, Eltem (1996) showed that fresh whole black olives, fresh damaged black olives, and whole black olive paste inhibited or greatly reduced the production of aflatoxins by toxinogenic strains of A. flavus and A. parasiticus isolated from naturally fermented black olives. Although these studies concurred to suggest that olives, especially the black, are not suitable substrates for the production of mycotoxins at hazardous levels, the occurrence of various mycotoxins in table olives in North African countries and elsewhere has extensively been documented (Table 10). A study carried out by Tantaoui-Elaraki and Le Tutour (1985) revealed the presence of the AFB1 and OTA at appreciable amounts in commercial samples of Moroccan table black olives in spite of the fact that these mycotoxins were shown not to be produced when black olives were artificially contaminated with selected strains of toxinogenic molds. Also, a survey carried out by Zinedine and others (2004) revealed that 36% of Moroccan table black olives were contaminated with OTA at levels ranging between 0.62 and 4.8 μg/kg with an average concentration of 1.43 μg/kg. Furthermore, OTA and AFB have been detected in 100% of sampled Moroccan dry-salted olives at levels exceeding 0.65 and 0.5 μg/kg, respectively (El Adlouni and others 2006). Therefore, the occurrence of toxigenic molds and the frequent detection of mycotoxins in traditionally processed black olives rank these products among the commodities of high risk to public health in North African countries. Such a risk is even greater when more than one mycotoxin is present in table olives, as was demonstrated by El Adlouni and others (2006) who showed the concomitant presence of OTA, CIT, and AFB in dry-salted commercial Moroccan Greek-style black olives. Furthermore, the prevalence of chronic nephropathy diseases in Tunisia has been correlated to the consumption of various foods contaminated with OTA, among which black table olives were incriminated, as abnormally high levels of this mycotoxin (up to 46.83 mg/kg) were found in this product (Maaroufi and others 1995a,1995b).
The decrease in moisture content to about 14% (Canellas and others 1993; Karathanos 1994) and the subsequent increase in sugar concentration in fruits during drying resulted in an almost selective medium for xerotolerant molds, among which members of the Aspergillus section Nigri have been shown to predominate. These black aspergilli are particularly severe and widespread in grapes of the warmer areas of the Mediterranean basin including North African countries (Battilani and others 2008). Strains of A. niger aggregate and A. carbonarius have been shown to be the main contaminants of Moroccan grapes and to produce high amounts of OTA (0.24 and 0.22 μg/g, respectively) at optimal environmental conditions (Selouane and others 2009). A similar situation is expected in the other grape-producing North African countries including Algeria, Tunisia, and Egypt, despite the lack of reports, as these countries share the same climatic conditions as well as sociocultural conditions. In addition to A. niger aggregate and A. carbonarius, A. aculeatus has also been found to predominate during the entire drying process of grapes; from the fresh fruits to the fully dried raisins (Leong and others 2003). However, this study showed that among the 3 fungal species, only A. carbonarius was able to produce OTA in vitro as detected by the emission of fluorescence under UV light upon cultivation on coconut cream agar plates. The incidence of occurrence of AFs and OTA in Moroccan commercial raisins and dried figs has been reported to be 30% and 65%, respectively, and the levels ranged between 0.03 and 1.42 μg/kg, with an average of 0.33 μg/kg in dried figs, and between 0.05 and 4.95 μg/kg with an average of 0.96 μg/kg in dried raisins (Zinedine and others 2007b). In Egypt, OTA was detected in figs at levels varying from 60 to 120 μg/kg, while raisin samples have proven to be free from AFs (B1, B2, G1, and G2), CIT, ochratoxins (OT), patulin (PAT), sterigmatocystin (STG), diacetoxyscirpenol (DAX), T-2 toxin, and zearalenone (ZEA). In another survey on 100 samples for the occurrence of mycotoxins, Egyptian raisins were found to be contaminated at very high levels of AFB1 (300 μg/kg) and OTA (250 μg/kg), though at the low frequencies of 2% and 1%, respectively (Youssef and others 2000). From the above data, it is clear that the occurrence of mycotoxins in fruits and vegetables presents a real threat to consumers’ health in these countries, as well as a serious limitation for the export of local produce toward developed countries, especially the traditional economic partners of the European Union (EU) and North America. The latter countries are in the process of harmonizing their regulations on mycotoxins in foods and feeds with a clear tendency to be more restrictive. On the other hand, there are presently no specific regulations on mycotoxins in North African countries, which is expected to hamper, on a medium or long run, the export of vegetable products from these countries to Europe and other partners from developed countries. It is worth mentioning that Morocco, Algeria, Tunisia, and Egypt are among the main suppliers of fruits and vegetables to Europe, which represents a major income to those North African countries where agriculture is the essential economic activity, especially in Morocco and Tunisia. However, despite the lack of specific regulatory limits for mycotoxins in these countries, the problem of mycotoxins is well recognized from both public health and economic views. Projects for the tolerable limits of aflatoxins and ochratoxins in foods have been proposed in some North African countries, but they have not come into force yet (Zinedine and Mañes 2009). This is essentially due to the belief of stakeholders in these countries that the strict implementation of mycotoxin regulations will have limited effects in terms of health protection. The prevalence of subsistence farming and the continued practice of traditional technologies for the transformation of food products in poor sanitary conditions will certainly hinder any effort to enforce such regulations. Indeed, the mycotoxin issue in North Africa or other developing countries needs to be viewed in the overall context of local food safety, health, and agricultural practice issues.