Cauim is a non-alcoholic drink produced from different substrates including rice, cassava, maize, peanut, cottonseed, banana and pumpkin by the Tapirapé Indians from Mato Grosso, Brazil. The traditional production of cauim involves cooking the different raw materials for approximately 2 h and then cooling them to room temperature. At this time the inoculum is added, which consists of the juice obtained from the chewing of sweet potatoes by the women of the tribe and fermented for 24 to 48 h (Schwan et al., 2007).

Schwan et al. (2007) studied the yeast diversity using a culture-dependent approach of a non-commercial cauim made with cassava and rice. The beverage was prepared in a traditional manner. To prepare the inoculum, the women of the tribe chew the sweet potatoes and add the saliva and the chewed sweet potatoes to the cassava and rice mixture (previously ground and cooked); in this way, the microorganisms present in the saliva act as starters. The role of saliva is to provide amylase to start the process of hydrolysing starch to fermentable sugar (Elizaquível et al., 2015). They found seven different yeast species, with Candida tropicalis and Candida intermedia being the most abundant. In another work, Ramos et al. (2010) studied the microbial diversity of another non-commercial cauim made with peanut and rice using culture-dependent (D) and -independent (I) methods. They also found that the dominant yeast genera at the end of fermentation were Candida (I/D) along with Saccharomyces cerevisiae (I/D) and Kluyveromyces lactis (I/D). The authors point out that the yeasts present in cauim fermentation would increase its nutritional value since they would be involved in starch degradation, flavour production, rising of vitamin B levels and free amino acid production (Schwan et al., 2007).

Ramos et al. (2010) also studied the bacterial diversity and found that Lactiplantibacillus plantarum (I/D) and Limosilactobacillus fermentum (I/D) dominated in the final stages of cauim fermentation.

Tarubá is a beverage that is distinctive since a solid substrate is used for fermentation. The cassava roots are cleaned, crushed, pressed, sieved and toasted for 30 min, placed in a wood receptacle (gareira) and covered with Trema micrantha (L.) Blume and Musa spp. (banana) leaves, and allowed to ferment for 12 days. To obtain the final beverage, the fermented cassava is diluted in water and filtered (Figure 2; Ramos et al., 2015). Examination of the microbial composition of this traditionally made non-commercial tarubá by culture-dependent and independent methods (PCR-DGGE; Ramos et al., 2015) shows that the dominant yeast throughout the fermentation process and in the final beverage was Torulaspora delbrueckii. With respect to bacterial taxonomy the authors found that at the end of the fermentation and in the final product that the dominant species were L. plantarum, Leuconostoc mesenteroides and Bacillus subtilis. It was suggested by the authors of this study that interactions between these microbes were responsible for flavour development due to the protease activity and volatile compound production.

Another fermented drink, produced by the Jaruma people, an indigenous population of the Mato Grosso, Brazil, is yakupa, a yellowish, slightly acidic and bitter beverage. It is a non-alcoholic beverage drank by the entire population daily (Lima, 2006). The preparation begins by letting cassava soak in water for 2 or 3 days, producing a pre-fermentation of the substrate (puba). The puba is dried in the sun and crushed, water is added and the mixture is sifted and cooked adding grated sweet potato. The drink can be consumed immediately or it can be left to ferment for 24 to 48 h (Freire et al., 2014).

Freire et al. (2014) studied the microbiological dynamics of non-commercial yakupa fermented over 60 h, produced by the local Amerindian Yudjá–Pakaya tribe, which is located in the Xingu Indigenous Park (Mato Grosso, Brazil) by culture-dependent and -independent (PCR-DGGE) methods. Throughout the entire fermentation process the authors found that L. fermentum and S. cerevisiae were the dominant species using both methodologies.

Another non-alcoholic fermented beverage is calugi, which is produced by Javaé Brazilian Indians (Tocantins, Brazil) using different substrates such as rice, cassava and maize (Miguel et al., 2012). Miguel et al. (2012) characterised the microbiological diversity of non-commercial calugi produced from rice and maize using the mastication juice of sweet potato as an inoculum. The maize (previously peeled) and rice were cooked together for 2 h. Once the mixture had cooled, the inoculum was added and allowed to ferment at room temperature (30°C approx.) for 48 h.

Culture-independent analysis (PCR-DGGE) showed that Streptococcus salivarius was the most abundant bacteria and was present throughout the fermentation. S. cerevisiae was present during almost the entire process, and was the only yeast detected at 46 h of fermentation (Miguel et al., 2012).

The Amerindian TapirapéeTapi’itãwa tribe from Mato Grosso, Brazil prepares cottonseed and rice fermented beverages for local consumption. Ramos et al. (2011) studied the microbial diversity using both culture-dependent and -independent (DGGE-PCR) methods. The cottonseed and rice were soaked in water and crushed. After that, water was added and cooked for 30 min, and then left to ferment in an open vessel for 48 h at room temperature (approximately 30°C). The authors (Ramos et al., 2011) found that the most abundant genera in the beverage were LAB (57%), mainly L. plantarum and Lentilactobacillus spp. at the end of fermentation. The yeast species found in the final product were Candida parapsilosis and Clavispora lusitaniae. LAB were the principal microorganisms present at the end of the fermentation process.

Champu is a fermented alcoholic beverage with a sweet, sour taste consumed in the rural and urban areas of Colombia. The ingredients can include cereals (wheat, rye and maize), pineapple, or “lulo” (Solanum quitoense Lam.), “panela” (sugar cane tablets) syrup, clove, cinnamon and orange tree leaves. The first step in the preparation of champu is to soften maize grains by boiling them for several hours, before adding the other ingredients. These are then left to ferment for 24 to 48 h at a temperature of 12–15°C. The resulting alcoholic content in the beverage is between 2.5% and 4.2% with a pH between 3.5 and 4 (Osorio-Cadavid et al., 2008).

Osorio-Cadavid et al. (2008) studied the yeast diversity present in 20 different commercial champu through a culture-dependent analysis (Table 1). They found 8 different species of yeast, where Issatchenkia orientalis, S. cerevisiae and Pichia fermentans were the most abundant species. According to the authors (Osorio-Cadavid et al., 2008), the diversity of yeasts present may be a reflection of the raw material used. These species were also found in different fermented beverages made using the same substrates (Sefa-Dedeh et al., 1999; Wacher et al., 2000; Botes et al., 2007), in sour cassava (Lacerda et al., 2005) and coffee fermentations (Masoud et al., 2004).

Chicha is a traditional homemade fermented beverage produced by the native populations in South America (Figure 2). This alcoholic beverage (2%–12% v/v) is mainly made from maize but other raw materials such as rice, cassava, peanut, fruit, or carob can also be used. It is consumed during religious and agricultural festivities as well as during family and social events (Lorence-Quiñones et al., 1999; Delibes and Barragan Villena, 2008). In Peru, chicha is prepared in a traditional way by the indigenous population using local maize varieties, it is called chicha de jora and has a low alcoholic content (1%–3%; Delibes and Barragan Villena, 2008). Although the traditional fermentation process relied on saliva as an inoculum, this has increasingly changed to employ approaches to replace the amylase present in the saliva (Lorence-Quiñones et al., 1999).

Caxiri is a spontaneously fermented alcoholic beverage manufactured from cassava by the Yudja women of the Pakaya tribe from Brazil, consumed particularly in ceremonies and festivities (Lima, 2006). To prepare caxiri, cassava roots are fermented for 2 days in running water to soften the skin. Then, the cassava tubers are peeled, cut into small pieces, pressed, grated and roasted. The resulting flour is subsequently mixed with water and thereafter placed in barrels to initiate the fermentative process, which usually takes 24 to 48 h (Figure 2). During this fermentation, besides the degradation of cyanogenic compounds and formation of aromatic substances, softening of the roots occurs caused mainly by LAB (Kimaryo et al., 2000). Generally, caxiri is consumed up to 120 h after preparation (Santos et al., 2012).

The study of the microbiological composition of non-commercial caxiri prepared with casava and sweet potato with culture-dependent methods (Santos et al., 2012) revealed that the Bacillus genus (subtilis, cereus, amyloliquefaciens among others) were dominant since the beginning to the end of the fermentation processes. Regarding to the LAB, Pediococcus acidilactici was the only specie found at the end of the fermentation. At the beginning of fermentation, different specie of yeast were found, being with S. cerevisiae being the dominant species through all the process.

Other authors (Miguel et al., 2015) studied the microbial diversity present in a non-commercial caxiri prepared with cassava, maize and sweet potatoes. They found that S. cerevisiae was also the dominant yeast during the entire process, but LAB were the dominant bacteria, with L. fermentum and Lactobacillus helveticus being the dominant species, unlike what was found by Santos et al. (2012) where the Bacillus genus was dominant. For this study, the authors used culture-dependent methods and DGGE and PCR analysis.

The production of artisanal fermented beverages involves a consortium of microorganisms, mainly bacteria and yeasts. Throughout the fermentation process, the type of predominant microorganism may vary (Blandino et al., 2003).

Through the fermentation process, yeasts participate in the degradation of starch, in the development of flavour, increasing vitamin B levels and free amino acid production, with the consequent increase in nutritional value (Schwan et al., 2007). Saccharomyces cerevisiae to S. cerevisiae was the dominant yeast in several fermented beverages (cauim, yakupa, calugi, champu and chicha) This yeast is mainly associated with ethanol, CO₂ production and other aromatic components, such as esters, organic acids and carbonyl compounds (Janssens et al., 1992; Torner et al., 1992).

LAB are the main bacteria genus found in most fermented beverages. They have an important role due to their ability to metabolize different carbon sources (McDonald et al., 1990). The growth of these bacteria could favour the growth of yeasts due to the acid production. Likewise, the production of vitamins and nitrogenous compounds by yeasts stimulates the growth of LAB (Nout and Sarkar, 1999). Another essential role of LAB is the production of antimicrobials that, in addition to acid production, create unfavourable conditions for the growth of pathogens, toxigenic and spoilage organisms, improving the safety and stability of the final product (Miguel et al., 2012).

Bacteria belonging to the genus Bacillus were found in different fermented beverages (calugi, caxiri, cotton seed rice beverage, chicha and taruba). They may be involved in the secretion of various enzymes such as amylases, lipases, cellulases and proteases, which are essential in the fermentation process (Rusul and Yaacob, 1995; Amoa-Awua et al., 1997; Obilie et al., 2004).

Co-metabolism between the different microorganisms improves the characteristics of the final products through flavour development, improvement of food safety, reduction in cyanogen glucoside levels, among other activities (Damiani et al., 1996; Obilie et al., 2004; Kamda et al., 2015).

Kumis (or Koumiss) is a fermented dairy beverage originated in Central Asia and commonly produced and consumed today in Russia. However, here our focus is on Colombian kumis. While traditionally kumis is produced with mare’s milk, the fermented beverage produced in Colombia is based on raw whole cow’s milk (Chaves-López et al., 2011) This beverage is fermented in rural and urban areas of South West Colombia with production slightly different from other variations produced in other countries. It is very similar to kefir, a sour and fizzy milk-based fermented drink (Bourrie et al., 2016), except that a liquid starter culture is used by back-slopping from the previous ferment. The milk is allowed to ferment for 2–3 days, with sugar cane and cinnamon added, before serving as a low alcohol (1%–2%), creamy, sour and sparkling beverage (Chaves-López et al., 2011).

Research into Colombian kumis is limited, with much of the research to date originating from within one group, and the full microbial content of kumis from Colombia not yet elucidated. Kumis of other origins have been identified as having lactobacilli and yeasts as the main microbial components (Samat Kozhakhmetov et al., 2014). Analysis from Chaves-López et al. (2011); Chaves-López et al., (2012, 2014b) has investigated the composition of Colombian kumis, with a particular focus on the potential angiotensin I converting enzyme (ACE) inhibitory activity of the microbes identified. These investigations highlighted similar trends as was seen with kumis from other origins within commercially produced kumis, with LAB and yeast dominating the populations present with counts of 7.05–9.59 log CFU g⁻¹ and 6.26–8.65 log CGU g⁻¹, respectively. In addition to the yeast and LAB populations from kumis, Chaves-López et al. (2011) have identified enterococci as part of the main microbial composition of Colombian kumis with counts of 4.29–8.3 log CFU g⁻¹. Although the yeast and enterococcal populations of Colombian kumis were further characterised, a thorough investigation of LAB from the kumis was not completed (Chaves-López et al., 2014a). Amongst yeast species identified from Colombian kumis, Galactomyces geotrichum was most commonly isolated (22.58% of total yeasts isolated; Chaves-López et al., 2012). Several other yeast species were isolated from Colombian kumis (Table 1) with 21 different RAPD-PCR profiles identified from the 93 isolates (Chaves-López et al., 2012). Enterococci were also investigated in further detail from Colombian kumis with 72 isolates identified, with Enterococcus faecalis being the dominant species (Chaves-López et al., 2011).

From ACE1 inhibitory investigations on isolated strains from kumis, proteolytic activity was identified for several yeasts and enterococci investigated with these activities potentially contributing to the flavour profiles developed in the kumis. This activity may contribute to health benefits following consumption, although this was not fully determined through these investigations (Chaves-López et al., 2011, 2012). These strains could be utilised in the production of kumis, although some are likely better suited to such industrial processes than others (Chaves-López et al., 2014a).

As it is observed in many countries where cheese production occurs, there are a wide variety of cheeses available across South America, varying in their ingredients, production methods and fermentation processes. Despite the large variety of cheeses available within South America, only a small proportion of these have been investigated in relation to their microbial composition and only those are highlighted here. Although several immigrant methods are used to produce cheese in South America (with the exception of the indigenous Minas cheese) as these cheeses have been produced over many generations and slight variations to methods, they could now be considered indigenous.

Minas cheese is a semi-soft cheese produced in Minas Gerais state, Brazil with varieties named according to geographical region. Generally, this cheese is produced from raw cow’s milk, although occasionally pasteurised milks may be used, and ripened for up to 60 days, with ripening length varying by producer. The starter culture used in this cheese production (called pingo) is from the whey of the previous cheese production (Arcuri et al., 2013; Perin et al., 2017; Sant’Anna et al., 2019). Large variation has been identified within the microbial composition of Minas cheeses, with variation not only evident between geographical regions (Arcuri et al., 2013) but also between individual production farms (Perin et al., 2017). This variation is likely due to the combined impact of the use of a raw milk starter, which unsurprisingly varies between farms (Perin et al., 2017), and the use of the “back-slopping” approach through the use of the whey from the previous production. Although specific details are not given on production methodologies utilised for each sample it is possible that some variation may be introduced here also.

Studies of commercially produced Minas cheeses have also identified some variation between studies with Perin et al. (2017) identifying Lactobacillus species as predominant members, with variation between the species based on geographical origin, while Sant’Anna et al. (2019) identified Streptococcus as the genus to which most OTUs identified belonged. These differences may be due to factors such as ripening time variation between studies, which would strongly influence microbial composition, although total ripening time of cheeses investigated in the case of Perin et al. (2017) are not given. The variation observed here in dominant members would have an influence on the properties and characteristics of these cheeses (Ahmed et al., 2021), although this has not been specifically investigated in Minas cheeses. In addition, Sant’Anna et al. (2019) identified a variance in microbial species over time with Streptococcaceae and Planococcaceae present at all stages of the ripening process while Lactobacillaceae and Leuconostocaceae appear in the later stages of the ripening. Camargo et al. (2021) studied commercial Minas Gerais cheese from Entre Sierras region. Samples from 4 different producer were analysed, in the same way as Perin et al. (2017), and the authors found that LAB were the predominant bacteria amongst all the samples. Lactococcus was the main genus identified with. Lc. Lactis being present in all samples (Camargo et al., 2021). Microbial succession has also been investigated in other cheeses (Fuka et al., 2013; O’Sullivan et al., 2015), with the variation in species present influencing end products of biochemical events during ripening ultimately dictating the properties of the final cheese (McSweeney, 2004). Overall, a variety of factors seem to have an influence over the microbial composition of Minas cheese. The variance in the members identified may be associated with the variance in flavour and characteristics of these cheeses between regions.

Argentinian-produced cheeses are predominantly produced using milks other than cow’s milk, especially goat and sheep, using immigrant methods commonly used in cheese production around the world. These semi-hard cheeses tend to be produced from the raw milks with little standardisation in the process involved (Medina et al., 2001; Oliszewski et al., 2007). The investigation of one such commercial semi-hard ewe’s milk cheese produced without a starter culture and which was ripened at 20°C for 30 days identified the dominance of L. plantarum, making up 93% of isolates (Medina et al., 2001). The L. plantarum were associated with important contributions to the ripening stages and flavour development, being involved with a number of enzymatic activities which may be responsible for the characteristic flavour of this cheese (Medina et al., 2001). Examination of an ewe’s milk cheese commercially produced in Brazil through similar processes (i.e. raw milk and no starter) identified a high proportion of LAB throughout the ripening period (3 months) from 6.0 CFUg⁻¹ to 6.7 CFUg⁻¹, with L. plantarum also identified here, although full characterisation of all strains present was not completed (Nespolo and Brandelli, 2010). Investigations of enzymatic activities of isolates from this cheese highlighted Lc. Lactis spp. lactis and L. plantarum as important contributors to lysis activities contributing to flavour development.

Goat’s milk in Argentina is generally used for the production of fermented foods rather than for direct consumption. Investigations of raw goat’s milk cheeses have yielded varying results between studies, which is not surprising given the use of raw milk and the absence of starters, similar to the process of the ewe’s milk cheeses. Although the diversity of species appeared to be higher in goat’s milk rather than that in ewe’s milk cheeses, this may be as a consequence of the greater number of studies and the variation in methods used to identify species with respect to the former. Across goat’s milk studies the most dominant species also varied, being either L. plantarum, L. fermentum, or Lacticaseibacillus rhamnosus depending on the study (Oliszewski et al., 2007; Rivas et al., 2014). Together these studies have identified Lactobacillus and Enterococcus species as the major LAB in goats milk cheeses that contribute to the ripening processes and flavour development through several enzymatic activities (Oliszewski et al., 2007; Rivas et al., 2014).

Several efforts have been made to attempt to standardise the production of ewe’s and goat’s milk cheeses in Argentina (Oliszewski et al., 2008; Bergamini et al., 2010). Both mixed and single strain starters have been investigated including strains of Streptococcus thermophilus, Lactobacillus helveticus, Lactobacillus bulgaricus, L. plantarum, L. rhamnosus and L. casei. As expected, different outcomes were apparent depending on the strain(s) used, however there were broadly beneficial effects with respect to the prevention of spoilage and pathogenic bacteria (Oliszewski et al., 2008; Bergamini et al., 2010).

Overall, despite the large variety of cheeses available within South America only a small proportion of these have been investigated. Variation in microbial composition could be observed between products originating from different milks (cow’s, goat’s, or ewe’s), which is to be somewhat expected, although geographical region and production method variation could also be at play here. Cheeses investigated across studies are predominantly artisanally produced, with little consistency in methodologies (for example ripening times vary, when outlined, from 10 to 90 days), and so as such the microbial diversity observed here in species identified is expected, although some species including L. plantarum were found to be commonly identified within many cheeses which is in line with investigations of cheeses globally (Xanthopoulos et al., 2000; Belicová et al., 2013; Fernandes et al., 2017). This species is well characterised with respect to its probiotic potential and importance in food safety, making it an important member of these cheeses (Dinev et al., 2018; Khalil et al., 2022).

Suero costeño is an indigenous spontaneously fermented cow’s milk product produced in the Colombian Caribbean coast, with primary production occurring in rural areas in the departments of Antioquia, Bolívar., Córdoba and Sucre. This sour cream-like product is traditionally produced using raw cow’s milk with the final product having a slightly lumpy texture with an acidic and salty taste. Raw milk is traditionally added to calabashes (container made from a gourd), although plastic vessels can also be used, with the continuous use of the same vessels and the environment seeding the fermentation (Cueto et al., 2007; Motato et al., 2017; Tirado et al., 2017; Castillo-Solano et al., 2018). Although not typical in suero costeño production, some producers also utilize back-slopping (Motato et al., 2017). Fermentation occurs over a period of 12 h to 3 days, depending on the desired viscosity, at an ambient temperature (~30°C) and at a high humidity. The resultant fermented product is in 2 phases with the liquid whey removed and the creamier product retained and salted at 1–3% for consumption as suero costeño, most commonly as a dressing (Cueto et al., 2007; Motato et al., 2017; Tirado et al., 2017; Castillo-Solano et al., 2018). This product is most commonly artisinally produced as any industrial production, which commonly uses pasteurised milk, is not widely accepted due to the variance in taste and texture (Tirado et al., 2017).

Only a few studies have investigated the microbiological composition of suero costeño with variation presumably due to a combination of the impact of the raw milk used, spontaneous fermentation and short fermentation time. Counts of bacteria and yeasts varying from 9 × 10¹ to 1 × 10⁸ CFU ml⁻¹, with lower counts identified with one producer who utilised back-slopping (Cueto et al., 2007; Motato et al., 2017). Lower counts of coliforms and Listeria counts were particularly evident for cheeses made using back-slopping (Motato et al., 2017), although overall these bacteria were low across samples (Cueto et al., 2007; Motato et al., 2017). Although back-slopping is not routinely utilised for the production of suero costeño, this potentially indicates a possible benefit in its use to control undesirable microbial growth during fermentation.

Lactobacillus, Lactococcus, Leuconostoc, Streptococcus, Aeromonas and Acetobacter were the dominant genera within commercially produced suero costeño samples, with a variance observed in dominant members identified between producers, regions and even batches (Cueto et al., 2007; Motato et al., 2017). Other taxa were also observed at lower abundances or rarely amongst samples investigated (Table 1), with some variance also depending on methodologies used (Cueto et al., 2007; Motato et al., 2017). Samples from different manufacturers investigated by Motato et al. (2017) from within the same area (Planeta Rica) were more similar to each other than the samples produced in the other area investigated (Caucasia). Although LAB were identified as the predominant members of the final fermentation when both UHT and raw milk were used, the microbial succession varied with aerobic mesophilic bacteria or yeasts dominating at earlier time points with UHT and raw milk, respectively (Cueto et al., 2007). Although the final microbial populations were similar, the variance during the fermentations between inputs may result in a variance in taste and texture of the final product.

The variance of these inputs may offer a potential to alter suero costeño to consumers’ preferences through the variance of the starter substrate. Another way in which this might be achieved could be through the addition of substrates at the beginning of the fermentation. For instance, the addition of whey to a goat’s milk suero costeño improved overall texture of the product, offering a potential alternative to the production with cow’s milk (Tirado et al., 2017). More research may be needed into this area for the understanding of the role of each of the members present within suero costeño and how the alteration of substrate may vary the final product and consumer perception.

Although dairy consumption is comparatively lower in South American countries there are still several fermented dairy products produced and consumed. Cow’s milk, as is the case in many countries, is the main substrate used in these fermentations, although certain regions favor the use of ewes or goats’ milks. These fermented dairy products share several common microbial members with L. plantarum, L. paracasei, Lc. lactis and L. casei commonly found amongst the majority of fermented dairy products from South America investigated. Although methodologies vary between studies reducing the potential for accurate comparisons, some geographical patterns could also be noted, for example with the presence of Pichia kudriavzevii in Colombian cow’s milk products and L. buchneri and L. parabuchneri in Brazilian cow’s milk products. Overall, an abundance of further research is needed to more thoroughly cover the breadth of diary fermented foods available in South America fully.

Salamis and salami-type sausages are the most common South American fermented meat products. Dry-cured sausages have traditionally been produced by a combined process of fermentation and no-heat drying treatments using different types of meat with some variations observed across regions or countries. Goat and lamb meat are used in some countries such as Brazil and Argentina. Llama (Lama glama) and guanaco (Lama paco) meat have a high protein value and low levels of fat and cholesterol, and are widely used in the Andean highlands in fermented meat products (Cristofanelli et al., 2004; Mamani-Linares and Gallo, 2013). Most fermented meat products from South America are artisanally manufactured. In Argentina, the manufacture of dry-fermented sausages (“salame”) in the main production areas is 10% through industrial processes and 90% through local artisanal techniques (Canel et al., 2013).

LAB (mainly lactobacilli) and Gram-positive coagulase-negative cocci (Kocuria and Staphylococcus) are the main bacterial groups present during dry-cured sausage fermentation. LAB rapidly lower the matrix pH and secrete bacteriocins which improves the hygienic safety preventing the growth of spoilage and pathogenic microorganisms. This also contributes to texture and color development. Kocuria and Staphylococcus also influence the typical cured red color through the reduction of nitrate to nitrite (Toldrá et al., 2014).

Based on the results by several studies on llama dry-cured sausages and as expected, the environmental and production conditions have a remarkable influence on the product bacterial composition, e.g., the occurrence of predominant species selected during the process was found in pilot but not in many artisanal productions, in which the same bacteria present at start where equally present at the end (Cocconcelli et al., 2012). Several genera and species were found in diverse studies, which are listed in Table 1. As indicated in them, Lactobacillus sakei seems to be a robust LAB species that is very well adapted to the harsh matrix and process conditions, strains of which have the potential to be developed as starter cultures for fermented meat products.

Charqui or charque, from the Quechua language “ch’arki,” is a type of salted and sun-dried meat of intermediate moisture with a shelf-life of months at room temperature. The broader category of this fermented meat incorporates charqui meat itself and its derivatives, jerked beef and carne-de-sol (sun meat; Shimokomaki et al., 2016). Charqui meats are typical of the Andean region, of Argentina, Chile, Peru, Bolivia and also Brazil. It has been consumed since pre-Columbian times to preserve meat for long periods. Historically, guanaco or llama meat was used (Figure 2), but with the arrival of the Spaniards in America, charqui made from beef became more common (Fadda and Vignolo, 2014).

For modern production, deboned raw meat muscles are injected with a brine solution, coated with coarse marine salt and piled up to let juices drain out. The piles are inverted every 24 h for several days, and in the final step, the salted meat pieces are exposed to the sun on rails until water activity (a_w) decreases to 0.7–0.8. The process can take up to 10 days. Due to the low a_w, the proliferation of pathogenic bacteria is unlikely, but halophilic and halotolerant bacteria may be able to grow. These include both bacteria favouring the fermentative process as well as spoilage bacteria, such as Halobacterium cutirubrum, which causes changes in odour and the appearance of slime and red spots on the product surface, coming from the salt used in the process (Pinto et al., 2002; Youssef et al., 2007). The fermentative process carried out by autochthonous microorganisms, mainly Micrococcus spp., Pediococcus halophilus, Staphylococcus xylosus, Lactobacilli. and Streptococcus spp. is essential for the unique sensorial characteristics of charqui, proteolytic activity and production of bacteriocins (Pinto et al., 2002).

Tocosh, togosh, or fute is a traditional fermented food product that is still prepared in small communities in the highlands of Central Peruvian Andes and the south of Colombia. It is usually prepared with potatoes, but other plants such as maize, papalisa, oca, or arracacha are also used. However, in this review we will refer to the preparation of potato-based tocosh. The traditional process consists of digging a ground hole near a water spring in which large amounts of potatoes (normally discarded potatoes) are placed between straw layers (shicshi). Then, the pile is covered with rocks in order to prevent the tubers from being washed away by the slight water current; the potatoes are left to ferment in this running water for 1–12 months. During this time, potatoes undergo an enzymatic blackening or browning and are laid in a dry shaded area to drain the water. Once dried, the product can be stored for consumption or sale. However, most commonly, the product is sun-dried and ground to obtain a fine flour-type product that is used to prepare broths, stews and mazamorra de tocosh, which is prepared by boiling a mixture of tocosh, sugar, cinnamon, cloves and water, and then served cold (Ray and Didier, 2015; Jiménez et al., 2018).

Jiménez et al. (2018) investigated the LAB diversity associated with potato tocosh for the first time, using culturing and HTS approaches. The study was performed on freshly harvested potatoes, as well as 1- and 8-month fermented samples. Lactobacillus species (L. sakei, L. casei, Lactobacillus farciminis, Levilactobacillus. brevis, L. fermentum) and Leu. mesenteroides predominated on fresh potatoes, however Clostridium, Zymophilus and Prevotella genus were predominantly present in 1- and 8-months tocosh samples. Some LAB were examined in greater depth, with several strains of L. sakei and Leu. mesenteroides displaying amylase and phytate-degrading abilities, as well as EPS, group B vitamin (riboflavin and folate) production capacity and antibacterial activity (Jiménez et al., 2018). Yépez et al. (2017) also characterised the LAB strains from tocosh from the perspective of their antibacterial and antifungal activity against spoilage microbes, and the responsible antimicrobial compounds from the supernatant. They found that L. fermentum T3M3 and Leu. mesenteroides T1M3 have potential as antimicrobial producers by inhibiting bacterial foodborne pathogens and spoiler fungi directly as well as the activity of their cell-free supernatant activity against toxigenic fungi (Yépez et al., 2017). Yeasts and moulds were also counted, decreasing from 7.1 × 10³ (fresh potatoes) to 1.0 × 10² CFU/g (8-months sample).

Finally, yet another study, focused on tocosh from two varieties of potato, Yungay and Peruanita, the authors reported that viable mesophilic aerobic microorganisms increased from 1.8 × 10⁴ to 3.6 × 10⁷ cfu g⁻¹, fungi and yeasts from 2.4 × 10² to 4.4 × 10³ cfu g⁻¹ decreasing in the final stage to 5.0 × 10¹ cfu g⁻¹. LAB counts ranged from 1.5 × 10⁴ to 6.6 × 10⁸, with L. brevis-Lactobacillus acidophilus increasing from 1 × 10² to 3.6 × 10⁵ cfu g⁻¹ (Yábar and Reyes, 2017).

These few studies suggest that Lactobacillus and Levilactobacillus genera play an important role in tocosh fermentation. It is also probable that the organic acids and antimicrobials produced during the process are the responsible for the decrease in the fungal population, including toxigenic fungi, leading to a stabilization of the final product, as observed on many fermented foods.

Masa agria is a maize fermented dough made in a traditional way in Colombia. To prepare masa agria, the kernels of yellow maize are covered with water and buried in a hot place (35°C–40°C) for 4 days, before the kernels are washed, milled and dehydrated for 3 or 5 days. The quality and sensory characteristics depend of the fermentation conditions such as the fermentation time and the different microorganisms present (Chaves-López et al., 2014b). The masa agria produced is used to prepare different traditional products such as tamales, empanadas, carantanta among others (Chaves-Lopez et al., 2016).

Chaves-Lopez et al. (2016) studied six samples of commercial masa agria from different producers from Valle del Cauca (Colombia). They used HTS to study the microbial composition and found that the genus Lactobacillus (mainly L. fermentum) was most abundant, followed by Acetobacter and Weissella. The abundance of these microorganisms is also responsible for the low pH (3.1–3.7) registered. L. fermentum seems to be a key organism for fermentation of maize, making the large amounts of starch available to the overall community. The fermentation products (lactate, formate and ethanol) may also serve as carbon sources for organisms, such as yeasts.

Cassava derivatives form part of Brazilian ethnic foods and culture, such as, puba, tiquira spirit, carimã, farinha d’água and polvilho azedo. Moulds and bacteria (mainly LAB) are the first to colonise raw cassava, metabolising the pre-existing sugar. The traditional methods of preparing food and beverages made from cassava rely on enzymes produced by moulds and naturally present bacteria, which hydrolyse the starch, producing fermentable sugars, as well as the yeasts that ferment these sugars to ethanol.