International Journal of Food Microbiology 63 (2001) 189197 www.elsevier.

nl / locate / ijfoodmicro

The microbiology of South African traditional fermented milks

a, b a Elisabeth M. Beukes *, Bernie H. Bester , Johannes F. Mostert
a

ARC- Animal Nutrition and Animal Products Institute, Private Bag X2, Irene 0062, South Africa b Department of Food Science, University of Pretoria, Pretoria 0002, South Africa Received 22 May 2000; received in revised form 26 July 2000; accepted 3 August 2000

Abstract A total of 15 samples of traditional fermented milk were collected from individual households in South Africa and Namibia. Lactic acid bacteria dominated the microora of these samples, especially the genera Leuconostoc, Lactococcus and Lactobacillus. Other groups identied included pyogenic streptococci and enterococci. The dominant lactococci species was Lactococcus lactis subsp. lactis. Eighty-three percent of the leuconostoc isolates were identied as Leuconostoc mesenteroides subsp. dextranicum. Other species identied included Leuconostoc citreum, Leuconostoc lactis, Lactobacillus delbrueckii subsp. lactis and Lactobacillus plantarum. 2001 Elsevier Science B.V. All rights reserved.
Keywords: Traditional fermented milk; Lactic acid bacteria; Maas; Amasi

1. Introduction Many people throughout Africa enjoy soured milk products. In these products, the lactic acid bacteria perform an essential role in preserving a highly nutritious food product. Fermented milk products are also of great signicance for their therapeutic value, for alleviating lactose intolerance, social value and as a means of generating income. The peoples of South Africa used to ferment their milk in milk-sacks, calabashes, clay pots, stone jars and baskets (Fox, 1939; Quinn, 1959; Bryant, 1967; Fehr, 1968; Bohme, 1976; Moifatswane, 1995 National Cul*Corresponding author. Tel.: 1 27-12-672-9041; fax: 1 27-12665-1563. E-mail address: elbie@iapi.agric.za (E.M. Beukes).

tural History Museum, personal communication). The art of making these products was handed down from one generation to the next. The Xhosa and Zulu people used mainly calabashes to make amasi while the South-Sotho preferred clay pots to make their ma. Fox (1939) stated that clay pots gave a better avour to the fermented milk than calabashes. A calabash needed to be seeded with a microbial inoculum before it could be use for the production of fermented milk. According to a description of Bryant (1967), a calabash lled with fresh milk was covered and placed outside the hut. After coagulation the whey was drained through a hole in the bottom of the calabash. The hole was initially stoppered with a piece of stalk. The calabash was now ripened and lled with fresh milk again. Fermentation took place within 2 to 3 h whereafter the curd was poured out as

0168-1605 / 01 / $ see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0168-1605( 00 )00417-7

190

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197

snow-white lumps. The specic container used, as well as the environmental conditions, contributed to the gradual selection of specic micro-organisms that were responsible for the rich, full avours that cannot easily be imitated by modern dairy starter cultures. Although traditional fermented products still enjoy loyal following in rural communities in South Africa (Hughson, 1995), the traditional containers are increasingly being replaced with commercially available ones, especially plastic containers (Coetzee et al., 1996). There is no scientic information available on traditional fermented milk in South Africa. Modern socio-economic changes mean that some traditional technologies for the production of fermented foods will eventually be lost together with the associated micro-organisms. It is therefore imperative that traditional, indigenous products as well as the preservation and exploitation of the associated fermentative micro-organisms be investigated. The objectives of this study were to collect indigenous fermented milk samples in different rural areas, to determine the predominant microbial groups and to identify the lactic acid bacteria in the fermented milks.

2. Materials and methods

2.1. Collection of samples

Fifteen samples of indigenous fermented milk were obtained from individual households in rural areas in northern South Africa and Namibia. These samples were collected from July 1995 to September 1996. Seven samples were in clay pots, six in calabashes and two in plastic containers. On receipt the pH of the samples was measured and the microbiological analyses performed within 24 h. The clay pots and calabashes were covered with tinfoil and stored under refrigeration at 4 to 78C.

2.2. Isolation of micro-organisms

Ten millilitres of each sample was pipetted aseptically into 90 ml of quarter strength Ringers solution and mixed thoroughly. Serial dilutions (10 2 1 to 10 2 8 ) were made and 1 ml portions of the appropriate dilutions were pour-plated on the following media. (i) Plate count agar (Oxoid M325) (Oxoid,

Basingstoke, Hampshire, UK), incubated at 30618C for 7262 h for enumeration of total aerobic plate count. The total colony count was determined as described in the International Dairy Federation (1991) reference method (IDF 100 B: 1991). (ii) MRS agar (De Man et al., 1960) (Oxoid CM 361), incubated anaerobically for 4862 h at 42618C for enumeration of thermophilic lactobacilli and streptococci. MRS agar plates were also incubated anaerobically for 4862 h at 35618C for enumeration of mesophilic lactobacilli and leuconostocs. (iii) M17 agar (Terzaghi and Sandine, 1975) (Oxoid CM 785), incubated aerobically for 4862 h at 30618C for the enumeration of lactococci. (iv) Rogosa agar (Rogosa et al., 1951) (E. Merck, D-61 Darmstadt), incubated anaerobically for 4862 h at 35618C for enumeration of lactobacilli. (v) Violet red bile agar (Oxoid CM 107 with added MUG supplement BRO 71 E), incubated aerobically for 2462 h at 37618C for enumeration of coliforms and Escherichia coli. The supplement containing 4-methylumbelliferyl-BD-glucuronide (MUG) allowed the separate enumeration of Escherichia coli which contain glucuronidase activity. Plates were examined under long wave UV light (366 nm) for the presence of uorescing colonies. (vi) Milk agar (Harrigan and McCance, 1976) was used for detection of proteolytic bacteria. The plates were incubated for 2448 h at 258C. (vii) Citrate fermenting organisms were detected on the modied medium of Nickels and Leesment (Vogensen et al., 1987), incubated for 48 h at 258C. Anaerobic jars (Biolab and Oxoid) with gas generating kits (Oxoid BR 38B) were used for anaerobic incubation. Ten isolates were obtained randomly from the countable plates of Rogosa agar, M17 agar, MRS agar (incubated at 428C for thermophilic bacteria and at 358C for mesophilic bacteria). Isolates were cultivated in MRS broth (Oxoid CM359) at 258C. Purity was checked by streaking on MRS agar. Isolates were preserved by a modication of the method described by Joubert and Britz (1987) and cultures were stored at 2 208C.

2.3. Identication of the lactic acid bacteria to genus level

Gram-positive, catalase-negative isolates were assigned to a genus on the basis of key characteristics and tests indicated in Table 1. Microscopic appear-

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197

191

Table 1 Differential characteristics of lactic acid bacteria (Harrigan and McCance, 1976; Garvie, 1984; Hammes et al., 1992; Holzapfel and Schillinger, 1992; Teuber et al., 1992; Weiss, 1992; Axelsson, 1993)a
Characteristic Leuconostocs Streptococci Pyogenes Cell form Spherical but often lenticular Pairs and chains Viridans Spherical or ovoid Lactic Spherical to ovoid Mainly in pairs, short chains Spherical Enterococci Pediococci Lactobacilli Strepto b Thermo c Rods/Coccobacilli Beta d

Cellular arrangement

Chains and pairs

Pairs, tetrads, clusters, single cells are rare, no chains

Chain formation common

Growth at 108C at 458C at 158C NH 3 from arginine Gas from glucose Growth in 6.5% NaCl Reaction in litmus milk

1 2 1 2 1 6 Comparatively inactive. Few strains capable of producing acid. Very few strains capable of clotting the milk. No strains giving reduction

2 2

2 1

1 2

1 1

6 6

ND 6 1 2 2 6

ND 1 2 6 2 6

ND 6 6 6 1 6

1 2 2 No reduction of litmus before clotting of milk

2 2 2 No reduction of litmus before clotting of milk

6 2 2 ARC

1 2 1 ARC

6 2 6 Comparatively inactive. Rarely produce sufcient acid to cause clotting

Various reactions depending on the species

1 , positive; 2 , negative; 6, response varies between species; ARC, acid, reduction, clot; ND, no data. Streptobacterium. c Thermobacterium. d Betabacterium.
b

ances of 24 h old cultures were judged using Gramstained preparations (Gerhardt et al., 1981). Growth at 10, 15 and 458C in MRS broth was evaluated visually after 72 h incubation. Tests for presence of catalase, production of ammonia from arginine and production of CO 2 from Gibsons medium were carried out as described by Harrigan and McCance (1976). The salt tolerance test was done using MRS broth, containing 6.5% (m / v) NaCl with incubation for 4 days at 378C. All isolates were also tested for their action on chalk litmus milk (Oxoid CM45 with added calcium carbonate) (Harrigan and McCance, 1976). Kanamycin aesculin azide agar (Oxoid CM591) was used for the presumptive identication of enterococci. Nickels and Leesment medium as modied by Vogensen et al. (1987) was used to differentiate between citrate fermenters and non-citrate fermenters. The insoluble calcium citrate is hydrolyzed by the citrate fermenters, which form a clear zone around the colony.

2.4. Identication of the lactic acid bacteria to the species level

Arginine tetrazolium agar (Turner et al., 1963; Harrigan and McCance, 1976) was used to differentiate between Lactococcus lactis subsp. lactis and Lactococcus lactis subsp. cremoris. The scheme outlined by Villani et al. (1997) was used for the presumptive identication of the leuconostoc species. The method of Garvie (1984) was used to divide the leuconostoc isolates into dextran-producing and nondextran-producing isolates. Subsequent tests done on these two groups differed: dextran-producing isolates were tested for fermentation of arabinose, maltose, rafnose, galactose and for growth at 378C; nondextran-producing isolates were tested for fermentation of sucrose, fructose, galactose and trehalose. Growth at 158C and growth in MRS broth containing 6.5% (m / v) NaCl were also tested. The basal medium described by Garvie (1984) was used for all

192

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197

the carbohydrate fermentation tests. The individual sugars were prepared as 2% (m / v) solutions, ltersterilised using a 0.45 mm porosity lter (Millipore, Bedford, MA, USA) and 0.5 ml of the sterile ltrate was added to 5 ml of basal medium. Growth studies at 15 and 378C were done in MRS broth. Ten representative isolates were selected for identication to species level using the API 50 CH galleries and API 50 CHL medium (bioMerieux sa). Bacterial cells were inoculated into tubules of the API galleries according to the manufacturers instructions. The galleries were incubated at 308C and reactions were observed after 24 and 48 h. The APILAB PLUS database (bioMerieux sa) was used to interpret the results.

1995b). Staphylococcus aureus was detected using the reference method of the IDF (International Dairy Federation, 1990).

3. Results

3.1. pH of the samples

The pH of 11 of the samples ranged from 4.0 to 5.4 with an average of 4.6.

3.2. Enumeration of micro-organisms

Table 2 summarises the microbial counts obtained from traditional fermented milk samples. Mean counts on MRS agar (358C) and M17 agar were 7.7 3 10 8 and 7.05 3 10 8 cfu ml 2 1 , respectively, and were higher than the mean total plate count (5.5 3 10 8 cfu ml 2 1 ), indicating the predominance of lactic acid bacteria. These counts also exceeded counts obtained on Rogosa agar. The mean thermophilic count (428C / 48 h) on MRS agar (3.86 3 10 8 cfu ml 2 1 ) was less than the mean mesophilic count (358C / 48 h) (7.7 3 10 8 cfu ml 2 1 ) for calabash and clay pot samples. The presence of proteolytic organisms in high numbers (mean value for 12 samples: 1.89 3 10 8 cfu ml 2 1 ) contributed substantially to the total bacterial population of the samples tested. The

2.5. Detection of pathogens

All samples were tested for the presence of Salmonella, Staphylococcus aureus and Listeria monocytogenes. The IDF reference method (International Dairy Federation, 1995a) was used for presumptive detection of Salmonella species with the use of Brilliant Green agar, Bismuth Sulphite agar and XLD (Xylose Lysine Desoxycholate) medium as selective solid media. Listeria selective medium (Oxford formulation CM856) as well as PALCAM Listeria selective agar (Oxoid CM877) were used for isolation and identication of presumptive Listeria monocytogenes (International Dairy Federation,

Table 2 Microbiological prole of samples from indigenous fermented milk in South Africa Medium Ranges of counts (cfu ml 2 1 ) for all samples (n 5 13) Total plate count agar MRS agar (428C) MRS agar (358C) M17 agar Rogosa agar Violet red bile agar Milk agar Nickels and Leesment medium 8.6 3 10 5 1.56 3 10 9 (n 5 12) 4.7 3 10 5 1.27 3 10 9 6.5 3 10 6 2.03 3 10 9 1.4 3 10 6 1.87 3 10 9 3.2 3 10 5 5.1 3 10 8 , 11.51 3 10 7 (n 5 12) 3 3 10 4 9 3 10 8 (n 5 12) , 18.4 3 10 8 (n 5 11) Mean counts (cfu ml 2 1 ) Clay pots (n 5 7) 9.43 3 10 8 (n 5 6) 7.65 3 10 8 1.28 3 10 9 1.2 3 10 9 2.47 3 10 8 3.05 3 10 6 (n 5 6) 3.62 3 10 8 1.47 3 10 8 Calabashes (n 5 6) 1.58 3 10 8 7.6 3 10 6 2.6 3 10 8 2.1 3 10 8 1.53 3 10 8 5.8 3 10 4 1.67 3 10 7 (n 5 5) 1.12 3 10 8 (n 5 4)

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197

193

highest coliform counts (3.2 3 10 6 and 1.5 3 10 7 cfu ml 2 1 ) were obtained from two clay pot samples. An important nding was the presence of Escherichia coli in three clay pot samples with counts varying between 4 3 10 3 , 7 3 10 3 and 19 3 10 3 ml 2 1 .

distribution of the 366 bacteria identied from traditional fermented milk in South Africa.

3.4. Identication of the lactic acid bacteria to species level

All the lactococci isolates (a total of 103 isolates) belonged to Lactococcus lactis subsp. lactis. Eightysix isolates from a total of 104 Leuconostoc isolates produced dextran from sucrose. With the exception of two isolates, all of these isolates fermented maltose and galactose but not arabinose and rafnose, and grew at 378C. According to the scheme of Villani et al. (1997), Leuconostoc mesenteroides subsp. dextranicum and Leuconostoc carnosum are the only two species that form dextran from sucrose and do not form acid from arabinose. Since most strains of Leuconostoc carnosum did not grow at 378C (Dellaglio et al., 1995) the above-mentioned isolates were presumptively identied as Leuconostoc mesenteroides subsp. dextranicum. Eighteen of the Leuconostoc isolates were non-dextran-producing isolates. All of these formed acid from sucrose, fructose, galactose and trehalose and were able to grow at 158C and in the presence of 6.5% NaCl. According to the identication scheme these isolates could either be Leuconostoc paramesenteroides or Leuconostoc mesenteroides subsp. mesenteroides. From the 10 isolates identied with the API 50 CH identication system, three belonged to the species Lactobacillus plantarum, two belonged to Leuconostoc citreum, two belonged to Leuconostoc lactis, one belonged to Lactobacillus delbrueckii subsp. lactis, and one belonged to Leuconostoc

3.3. Identication of lactic acid bacteria to genus level

The greater part of the total number of isolates was Gram-positive and catalase-negative. A total of 366 isolates isolated from seven clay pots and two calabashes could be identied and were divided into ve genera: Lactococcus, Leuconostoc, Lactobacillus, Enterococcus and Streptococcus. Some heterofermentative lactobacilli grew as coccobacilli and were not easy to distinguish from leuconostocs. Twenty-nine isolates were assigned to the Leuconostoc / Betabacterium group, which means that they may either belong to the genus Leuconostoc or Lactobacillus. Eighty-four of the isolates were unmistakably rod-shaped and could easily be identied as Streptobacterium (57 isolates) or Betabacterium (27 isolates) (Harrigan and McCance, 1976). The ability to produce gas from Gibsons milk (Garvie, 1984) was an important characteristic for distinguishing the leuconostocs. One hundred and thirty isolates belonged to the genus Leuconostoc, 103 isolates to Lactococcus, 13 isolates to Streptococcus and seven isolates to Enterococcus. Members of the genus Lactococcus dominated in two clay pot samples, while Leuconostoc species prevailed in the calabash samples. Fig. 1 illustrates the percentage

Fig. 1. Identity of 336 bacteria isolated from indigenous traditional South African fermented milks.

194

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197

dextranicum. One of the isolates could not be identied.

3.5. Detection of pathogens

Staphylococcus aureus was isolated from fermented milk produced in a plastic container, clay pot and a calabash gourd. Salmonella species and Listeria monocytogenes were not detected in any of the samples.

4. Discussion The lactic acid bacteria predominated the microbial population and numbers between 4.7 3 10 5 and 2.03 3 10 9 were recorded with mean values of ca. 10 8 cfu ml 2 1 on MRS, M17 and Rogosa agars. The counts compared favourably with ndings of similar studies on fermented milks by other workers. According to Stadhouders (1975) the numbers of lactococci may easily reach 10 9 viable units per gram of sour milk. The dominance of mesophilic bacteria may be explained by the fact that 12 of the 15 samples were collected in the cooler months of May, July and August and the ambient temperatures at which the natural fermentation of the tested samples took place probably favoured proliferation of mesophilic bacteria. The higher counts on MRS agar (incubated at 358C) and M17 agar compared to counts on Rogosa agar may be explained by the fact that MRS and M17 agars are elective while Rogosa agar is selective (Reuter, 1985). The coliforms along with the yeasts formed the minority groups. The high coliform count in some of the samples was alarming when considering the South African health regulation that states that no person shall sell for consumption raw milk that has become sour which contains more than 50 coliform bacteria ml 2 1 of the product (South Africa, 1997). According to Hamama (1992), Moroccan traditional fermented dairy products like Lben and Jben showed high counts of indicator micro-organisms (e.g., coliforms, enterococci) and pathogens such as Salmonella spp., Yersinia enterocolitica, Listeria monocytogenes and enterotoxigenic Staphylococcus aureus. The methods of production of the various

traditional foods are usually primitive, compared to modern ways of food preparation (Dirar, 1997). Major risk enhancing factors are the use of contaminated raw materials, lack of pasteurisation, use of poorly controlled natural fermentations and inadequate storage and maturation conditions (Nout, 1994). The fact that no Lactococcus lactis subsp. lactis biovar. diacetylactis and Lactococcus lactis subsp. cremoris could be identied in this study was disappointing from both an ecological and practical point of view. All the Lactococcus isolates belonged to Lactococcus lactis subsp. lactis. According to other reports, Lactococcus lactis subsp. lactis was more frequently isolated than Lactococcus lactis subsp. cremoris from raw milk samples (Moreno and Busani, 1990), raw milk cheeses (Centeno et al., 1996), Raib (Morocco) (Hamama, 1992) and Dahi and buttermilk samples from India (PadmanabhaReddy et al., 1994). According to Holler and Steele (1995), Lactococcus lactis subsp. cremoris was isolated only rarely from natural sources. According to Crow et al. (1993) and Weerkamp et al. (1996), lactococci isolated from natural sources were usually identied as Lactococcus lactis subsp. lactis, whereas the phenotype Lactococcus lactis subsp. cremoris, which is common in industrial mixed-strain starter cultures, was isolated only rarely. The natural habitat of Lactococcus lactis subsp. cremoris remains uncertain (Salama et al., 1995). From 21 isolates identied form Amazi, a fermented milk in Zimbabwe, Mutukumira (1996) found that ve isolates belonged to Lactococcus lactis subsp. lactis and four to Lactococcus lactis subsp. lactis biovar. diacetylactis. The fact that no Lactococcus lactis subsp. lactis biovar. diacetylactis could be detected in this study may be explained by the fact that only phenotypical tests were used for identication. Leuconostoc mesenteroides subsp. cremoris was also not encountered in this study. This may be explained by the fact that the characteristic of citrate metabolism is encoded on plasmid DNA, which may be lost in some strains (Cogan, 1985). Studies on 182 representative strains of lactic acid bacteria associated with raw milk in Brazil showed Leuconostoc mesenteroides subsp. cremoris as a minor group, representing only 1.1% of the total population (Antunes and De Oliveira, 1986, cited by Holzapfel and Schillinger, 1992).

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197

195

In the present study, three isolates were assigned to the species Lactobacillus plantarum. Of the identied 21 isolates from naturally fermented milk in Zimbabwe (Mutukumira, 1996), three were identied as Lactobacillus plantarum. From a total of 100 isolates from fermented milk in Northern Tanzania, Isono et al. (1994) identied four as Lactobacillus plantarum. From cultured milk in Cameroon, Jiwoua and Milliere (1990) identied 47 out of 426 isolates as Lactobacillus plantarum. According to Daeschel et al. (1987) (cited by Olasupo et al., 1997), Lactobacillus plantarum is known to be commonly associated with plants. Thus, in studies on the occurrence of lactic acid bacteria, Lactobacillus plantarum constituted the highest number of Lactobacillus species isolated from fermented plant materials (Olukoya et al., 1993; Olasupo et al., 1997). Leuconostoc lactis was one of the main species recovered from Jben, a traditional soft cheese from Morocco (Hamama, 1992). Only two strains were identied as Leuconostoc lactis in our study. From the total of 72 leuconostoc isolates obtained from a cheese), 31 raw cows milk cheese in Spain (Arzua were identied as Leuconostoc mesenteroides, 18 as Leuconostoc dextranicum, 22 as Leuconostoc paramesenteroides and one as Leuconostoc lactis (Centeno et al., 1996). The isolation of Leuconostoc citreum from dairy products has not been reported frequently by researchers. It was, however, isolated from Afuegal Pita cheese (Cuesta et al., 1996). Two of the 10 isolates in our study, which were characterised by the API 50 CH system, were Leuconostoc citreum. In a parallel study on the same traditional fermented milk samples, it was found that the yeast counts (mean value: 4.1 3 10 6 cfu ml 2 1 ) were noticeably lower than the counts for lactic acid bacteria (Loretan, 1999). Other researchers investigating fermented milk products reported varying amounts of yeasts and moulds present in those products. The presence of yeasts may be inuenced by the age of the product as well as the containers and processing methods used. In the production of Iria ri Matii in Kenya, glowing splints of wood are used to scratch the inside of the gourd. This process is repeated until the inside is smooth and even. The low counts of yeasts reported by Kimonye and Robinson (1991) may be explained by the presence

of inhibitory components of the smoke used in the treatment of the gourds. Hosono et al. (1989) found an appreciable number of yeasts (1.1 3 10 7 ml 2 1 ) in samples of Dadih, an Indonesian fermented milk, which is made by pouring buffalo milk into fresh bamboo tubes and capping them with banana leaves.

5. Conclusion In this study on traditional fermented milk in South Africa it was established that the method of preparing cultured milk in traditional containers such as calabashes and clay pots has diminished and is nowadays probably only practised in the most remote rural areas of the country. Modern containers that are readily available are replacing the calabashes and clay pots of which the fabrication and preparation is a dying art. The availability of milk is also determining the production of fermented milk products. The microbiological composition of the lactic acid bacteria found in the samples of traditional fermented milk products coincided with that of commercial mesophilic starter cultures with regard to the dominance of lactococci and leuconostocs. However, the industrially important species of Lactococcus lactis subsp. cremoris, Leuconostoc mesenteroides subsp. cremoris and Lactococcus lactis subsp. diacetylactis were not encountered in the natural products. Traditional fermented milk has successfully been upgraded in South Africa since the 1980s to large-scale industrial production in the form of Maas and Inkomasi (Keller and Jordaan, 1990). Commercial mesophilic cultures containing Lactococcus lactis subsp. lactis, Lactococcus lactis subsp. cremoris and Leuconostoc mesenteroides subsp. cremoris are being used in the production of these products. In addition to the production of safer foods, industrialisation has further advantages such as standardised products (Reilly and Westby, 1997), reduced processing times, increase in the production of traditional dairy products as well as better distribution and marketing (Hamama, 1992). According to Hughson (1995), traditional fermented milk has been commercialised in South Africa with an estimated production of 60 000 kilolitres in 1995. She also stated that there was a scarcity of new product development for fermented dairy products targeted at the black consumer mar-

196

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197 Fehr, W., 1968. Ludwig Albertis Account of the Tribal Life and Customs of the Xhosa in 1807. A.A. Balkema, Cape Town, 22 pp. Fox, F.W., 1939. Some Bantu recipes from the Eastern Cape Province. Bantu Studies XIII (1), 6574. Garvie, E.I., 1984. Separation of species of the genus Leuconostoc and differentiation of the leuconostocs from other lactic acid bacteria. In: Bergan, T. (Ed.). Methods in Microbiology, Vol. 16. Academic Press, London, pp. 147178. Gerhardt, P., Murray, R.G.E., Costilow, R.N., Nester, E.W., Wood, W.A., Krieg, N.R., Phillips, G.B., 1981. Manual of Methods For General Bacteriology. American Society for Microbiology, Washington, DC. Hamama, A., 1992. Moroccan traditional fermented dairy products. In: Ruskin, F.R. (Ed.), Applications of Biotechnology To Traditional Fermented Foods. National Academy Press, Washington, DC, pp. 7579. Hammes, W.P., Weiss, N., Holzapfel, W.H., 1992. The genera Lactobacillus and Carnobacterium. In: Barlows, A., Truper, H.G., Dworkin, M., Harder, W., Schleifer, K.-H. (Eds.). The Prokaryotes, Vol. 2. Springer, Berlin, pp. 15341593. Harrigan, W.F., McCance, M.E., 1976. Laboratory Methods in Food and Dairy Microbiology. Academic Press, London. Holler, B.J., Steele, J.L., 1995. Characterization of lactococci other than Lactococcus lactis for possible use as starter cultures. Int. Dairy J. 5, 275289. Holzapfel, W.H., Schillinger, U., 1992. The genus Leuconostoc. In: Barlows, A., Truper, H.G., Dworkin, M., Harder, W., Schleifer, K.-H. (Eds.). The Prokaryotes, Vol. 2. Springer, Berlin, pp. 15091534. Hosono, A., Wardojo, R., Otani, H., 1989. Microbial ora in Dadih, a traditional fermented milk in Indonesia. LebensmWiss. Technol. 22, 2024. Hughson, L., 1995. Dairy products and the black consumer. Food Rev. 22, 3133. International Dairy Federation, 1990. Milk and Milk-based Products Enumeration of Staphylococcus aureus (IDF Standard 145: 1990). IDF, Brussels. International Dairy Federation, 1991. Milk and Milk Products Enumeration of Microorganisms. Colony Count at 308C (IDF Standard 100b: 1991). IDF, Brussels. International Dairy Federation, 1995a. Milk and Milk Products Detection of Salmonella (IDF Standard 93B: 1995). IDF, Brussels. International Dairy Federation, 1995b. Milk and Milk Products Detection of Listeria monocytogenes (IDF Standard 143A: 1995). IDF, Brussels. Isono, Y., Shingu, I., Shimizu, S., 1994. Identication and characteristics of lactic acid bacteria from Masai fermented milk in Northern Tanzania. Biosci. Biotechnol. Biochem. 58, 660664. Jiwoua, C., Milliere, J.B., 1990. Lactic ora and enterococci in cultured milk (Pindidam) manufacture in Adamaoua, Cameroon. Lait 70, 475486, abstract. Joubert, W.A., Britz, T.J., 1987. A simple and inexpensive method for the long-term preservation of microbial cultures. J. Microbiol. Methods 7, 7376. Keller, J.J., Jordaan, I., 1990. Fermented milks for the South African market. S. Afr. J. Dairy Sci. 22, 4749. Kimonye, J.M., Robinson, R.K., 1991. Iria ri Matii a traditional

ket. Ideas for new products may come as a development from existing ones or from re-discovering old ones (Marshall, 1986). The knowledge behind the production of traditional products has dwindled and in time will probably be lost forever. Gleaning of indigenous knowledge is not only meaningful for future developments, but also gives insight into the importance of this highly appreciated product for their consumers.

Acknowledgements The authors wish to thank the Agricultural Research Council for the nancial support of this project.

References
Axelsson, L.T., 1993. Lactic acid bacteria: classication and physiology. In: Salminen, S., Von Wright, A. (Eds.), Lactic Acid Bacteria. Marcel Dekker, New York, pp. 164. Bohme, H.E., 1976. Some Nguni crafts: calabashes. Ann. S. Afr. Museum 70, 915. Bryant, A.T., 1967. The Zulu People As They Were Before The White Man Came, 2nd Edition. Shuter and Shooter, Pietermaritzburg, 264 pp. Centeno, J.A., Cepeda, A.A., Rodriguez-Otero, J.L., 1996. Lactic acid bacteria isolated from Arzua cows milk. Int. Dairy J. 6, 6578. Coetzee, M.J., Gordeuk, V.R., Barnard, H.C., Stassen, A.F., De Kock, A., 1996. Letter to the editor: iron overload in a Venda community: a pilot study. S. Afr. J. Food Sci. Nutr. 8, 114 115. Cogan, T.M., 1985. The Leuconostoc : milk products. In: Gilliland, S.E. (Ed.), Bacterial Starter Cultures For Foods. CRC Press, Boca Raton, FL, pp. 2540. Crow, V.C., Coolbear, T., Holland, R., Pritchard, G.G., Martley, F.G., 1993. Starters as nishers: starter properties relevant to cheese ripening. Int. Dairy J. 3, 423460. Cuesta, P., Fernandez-Garcia, E., Gonzalez-De-Llano, D., Montilla, A., Rodriguez, A., 1996. Evolution of the microbiological and biochemical characteristics of Afuegal Pitu cheese during ripening. J. Dairy Sci. 79, 16931698. De Man, J.C., Rogosa, M., Sharpe, M.E., 1960. A medium for the cultivation of lactobacilli. J. Appl. Bacteriol. 23, 130135. Dellaglio, F., Dicks, L.M.T., Torriani, S., 1995. The genus Leuconostoc. In: Wood, B.J.B., Holzapfel, W.H. (Eds.), The Genera of the Lactic Acid Bacteria. Blackie Academic and Professional, London, pp. 235278. Dirar, H.A., 1997. Introduction. In: Dirar, H.A. (Ed.), Food Processing Technologies For Africa. United Nations Industrial Development Organization, Vienna, pp. 110.

E.M. Beukes et al. / International Journal of Food Microbiology 63 (2001) 189 197 fermented milk from Kenya. Dairy Ind. Int. 56, 3435, abstract. Loretan, T., 1999. The diversity and technological properties of yeast from indigenous traditional South African fermented milks. M.Sc. Thesis, Department of Microbiology and Biochemistry, University of the Free State, Bloemfontein. Marshall, V.M.E., 1986. The microora and production of fermented milks. Prog. Ind. Microbiol. 23, 144. Moreno, I., Busani, S.F.B., 1990. Characterization of lactococci isolated from raw milk and commercial lactic starters. Coletanea do Instituto de Technologia de Alimentos 20, 4450, abstract. Mutukumira, A.N., 1996. Investigation of some prospects for the development of starter cultures for industrial production of traditional fermented milk in Zimbabwe. Doctor Scientiarum Thesis, Department of Food Science, Agricultural University of Norway. Nout, M.J.R., 1994. Fermented foods and food safety. Food Res. Int. 27, 291298. Olasupo, N.A., Olukoya, D.K., Odunfa, S.A., 1997. Identication of Lactobacillus species associated with selected African fermented foods. Z. Naturforsch. 52, 105108. Olukoya, D.K., Ebigwei, S.I., Adebanwo, O.O., Osiyemi, F.O., 1993. Plasmid proles and antibiotic susceptibility patterns of Lactobacillus isolated from fermented foods in Nigeria. Food Microbiol. 10, 279285. Padmanabha-Reddy, V., Habibulla-Khad, M.M., Purushothaman, V., 1994. Plasmid linked starter characteristics in lactococci isolated from dahi and buttermilk. Cult. Dairy Prod. J. 29, 2526, 2830. Quinn, P.J., 1959. Foods and Feeding Habits of the Pedi. Witwatersrand University Press, Johannesburg, 132 pp. Reilly, A., Westby, A., 1997. Food fermentation Traditional low-cost food preservation systems. In: Dirar, H.A. (Ed.), Food Processing Technologies For Africa. United Nations Industrial Development Organization, Vienna, pp. 6978. Reuter, G., 1985. Elective and selective media for lactic acid bacteria. Int. J. Food Microbiol. 2, 5568. Rogosa, M., Mitchell, J.A., Wiseman, R.F., 1951. A selective medium for the isolation of oral and fecal lactobacilli. J. Bacteriol. 62, 132133.

197

Salama, S.S., Musaja-Jeknic, T., Sandine, W.E., Giovannoni, S.J., 1995. An ecological study of lactic acid bacteria: isolation of new strains of Lactococcus including Lactococcus lactis subspecies cremoris. J. Dairy Sci. 78, 10041017. South Africa, 1997. Foodstuffs, Cosmetics and Disinfectants Act (Act 54 of 1972). Proclamation No. R1555. Regulations relating to milk and dairy products. Government Gazette Vol. 389, 21 November 1997. Stadhouders, J., 1975. Microbes in milk and dairy products. An ecological approach. Neth. Milk Dairy J. 29, 104126. Terzaghi, B.E., Sandine, W.E., 1975. Improved medium for lactic streptococci and their bacteriophages. Appl. Microbiol. 29, 807813. Teuber, N., Geis, A., Neve, H., 1992. The Genus Lactococcus. In: Barlows, A., Truper, H.G., Dworkin, M., Harder, W., Schleifer, K.-H. (Eds.). The Prokaryotes, Vol. 2. Springer, Berlin, pp. 14821501. Turner, N., Sandine, W.E., Elliker, P.R., Day, E.A., 1963. Use of tetrazolium dyes in an agar medium for differentiation of Streptococcus lactis and Streptococcus cremoris. J. Dairy Sci. 46, 380385. Villani, F., Moschetti, G., Blaiotta, G., Coppola, S., 1997. Characterization of strains of Leuconostoc mesenteroides by analysis of soluble whole-cell protein pattern. DNA ngerprinting and restriction of ribosomal DNA. J. Appl. Bacteriol. 82, 578588. Vogensen, F.K., Karst, T., Larsen, J.J., Kringelum, B., Ellekaer, D., Waagner Nielsen, E., 1987. Improved direct differentiation between Leuconostoc cremoris, Streptococcus lactis subsp. diacetylactis and Streptococcus cremoris / Streptococcus lactis on agar. Milchwissenschaft 42, 646648. Weerkamp, A.H., Klijn, N., Neeter, R., Smit, G., 1996. Properties of mesophilic lactic acid bacteria from raw milk and naturally fermented raw milk products. Neth. Milk Dairy J. 50, 319332. Weiss, N., 1992. The Genera Pediococcus and Aerococcus. In: Barlows, A., Truper, H.G., Dworkin, M., Harder, W., Schleifer, K.-H. (Eds.). The Prokaryotes, Vol. 2. Springer, Berlin, pp. 15021507.