International Journal of Food Microbiology 43 (1998) 115–122

Microbial populations associated with commercially produced South
African sorghum beer as determined by conventional and
Petrifilm TM plating
Tracey-Lee Pattison, Ifigenia Geornaras*, Alexander von Holy
Department of Microbiology, University of the Witwatersrand, Private Bag 3, WITS, 2050, South Africa
Received 15 June 1998; accepted 15 June 1998

Abstract
Microbial populations of 46 commercially produced sorghum beer samples from retail outlets in Johannesburg, South
Africa, were enumerated and characterized. Aerobic plate counts, lactic acid bacteria counts and yeast counts were
performed by conventional and Petrifilm TM plating. Conventional methods yielded yeast counts of 7.84 log CFU / ml, lactic
acid bacteria counts of 6.44 log CFU / ml and aerobic plate counts of 5.96 log CFU / ml. In comparison, Petrifilm TM counts
were 7.85 log CFU / ml for yeasts, 5.31 log CFU / ml for lactic acid bacteria and 5.34 log CFU / ml for aerobic bacteria.
Characterization of 419 predominant bacterial isolates from Standard One Nutrient Agar, MRS Agar and corresponding
Petrifilm TM plates yielded 88.0% lactic acid bacteria, 8.4% Bacillus species, 2.9% Micrococcus species and 0.7% Gram
negative bacteria. Composition of predominant lactic acid bacteria populations from Standard One Nutrient Agar and both
types of Petrifilm TM plates showed marginal differences. Increased proportions of heterofermentative lactic acid bacteria
were, however, isolated from conventional MRS Agar compared to the modified Petrifilm TM product which represented the
equivalent to MRS Agar.  1998 Elsevier Science B.V. All rights reserved.
Keywords: Sorghum beer; Microbial populations; Petrifilm TM ; Lactic acid bacteria

1. Introduction
Fermentation is a low cost form of food processing and thus a significant component of the African
diet consists of fermented foods (Odunfa, 1985;
Cooke et al., 1987). One of these is sorghum beer
which is a soured fermented drink indigenous to
sub-Saharan Africa. It is made from malted sorghum
*Corresponding author: Fax: 1 27 11 339 7377; e-mail:
gina@gecko.biol.wits.ac.za

and unmalted maize (Odunfa, 1985) and differs
considerably from conventional clear lager type
beers (Haggblade and Holzapfel, 1989). It is opaque
and pinkish–brown in colour due to large quantities
of suspended particles and yeasts (Odunfa, 1985).
Being more viscous than lager beer, it resembles a
thin gruel with a sour, yoghurt-like taste, which
makes it refreshing even though it is normally
consumed warm (Haggblade and Holzapfel, 1989).
Due to its relatively low alcohol content, which
generally does not exceed 3%, and the large quantity

0168-1605 / 98 / $19.00  1998 Elsevier Science B.V. All rights reserved.
PII: S0168-1605( 98 )00103-2

Subsequent formation of lactic acid after the souring stage is arrested by boiling of the mixture which kills all microorganisms in the vegetative form (Haggblade and Holzapfel. 1989). 1989). sour taste. 1956). 1989). souring. Petrifilm TM was evaluated in conjunction with conventional methods to assess its suitability for sorghum beer quality control. The main cause of spoilage in sorghum beer is reportedly attributable to mesophilic. 1989). Lactic acid bacteria. Since alcoholic fermentation is carried out shortly before distribution to the retail trade. A typical flow diagram for the industrial production of South African sorghum beer is shown in Fig. can also contribute to the spoilage of sorghum beer (Haggblade and Holzapfel. Unlike clear beer production. boiling. however. 1996).5 of the mixture at the end of souring (Haggblade and Holzapfel. yoghurt-like taste favourable to consumers. 1989. 1989). Typical process flow illustrating commercial sorghum beer production. This study investigated microbial numbers and bacterial populations of sorghum beer sampled from retail outlets in the Johannesburg area. to allow the escape of carbon dioxide (Haggblade and Holzapfel. the bacterial spoilage populations in the product at the point of sale are not. Odunfa. and carbon dioxide (Haggblade and Holzapfel. The final sorghum beer product is not pasteurized. Heterofermentative lactobacilli also contribute to the unacceptability of the product by producing acetic acid or ethanol in addition to lactic acid. 1989). / International Journal of Food Microbiology 43 (1998) 115 – 122 of suspended solids. It is commonly stored at room temperature at the point of sale. Pattison et al. and thus sold in a microbiologically active state (Haggblade and Holzapfel. The lactic acid produced by the lactic acid bacteria during the souring stage is particularly desirable since it promotes the distinctive sour. the production of sorghum beer involves two fermentations: a lactic acid fermentation and an alcoholic fermentation. Tlhoaele. The lactic acid fermentation or souring stage of sorghum beer production confers the characteristic Fig. 1989). storage qualities and also assures consumer safety of the final beer product (Haggblade and Holzapfel. 3. resulting in a short shelf-life of 1–5 days (Haggblade and Holzapfel. the product is sold in a state of active yeast fermentation and is thus packaged into carton or plastic containers that have a small vent at the top. the metabolic activities of which result in a pH of ca. 1966. The brewing of indigenous sorghum beer involves malting. 1. 1985). mashing. 1989).116 T. . homofermentative lactobacilli which are commonly introduced in the final stages of the process (post second conversion. 1) and cause oversouring as a result of lactic acid production (Haggblade and Holzapfel. straining and alcoholic fermentation (Haggblade and Holzapfel. Fig. The alcoholic fermentation is the final step in the production of sorghum beer and is usually initiated by pitching of the wort with active dry brewer’s yeast consisting of a top fermenting strain of Saccharomyces cerevisiae (Novellie. 1963). many consumers consider this beer as much a food as a beverage (Novellie. 1989). While lactic acid bacteria associated with the souring stage of sorghum beer production are well characterized. The predominant bacterial group reportedly responsible for the lactic acid fermentation is Lactobacillus (van der Walt. 1.

St Louis. Cycloheximide (Sigma. UK). Microbiological and chemical analyses were performed on the day of sampling. / International Journal of Food Microbiology 43 (1998) 115 – 122 2. according to the manufacturer’s instructions (Table 1). 117 One ml of the above dilution series was also added to duplicate Petrifilm TM -Yeast and Mold plates (Table 1). The exact age of the samples was not known. b Cycloheximide (20mg / ml) added to diluent used for tenfold sample dilutions Petrifilm TM -Aerobic Plate Count plates b Modified-Petrifilm TM -Aerobic Plate Count plates b (with quadruple strength MRS Broth. Midrand. pH 3. but was estimated to range from 2 to 4 days.7 TM Petrifilm Aerobic plate count (PF-APC) Lactic acid bacteria count (PF-LABC) aerobic anaerobic a Yeast count (PF-YC) aerobic a Oxoid Anaerobic System.K. Basingstoke.2.1. Materials and methods 2. were purchased from two retail liquor stores in Johannesburg suburbs once a week over a 3 month period (April to June). Basingstoke. Sampling procedure Forty-six sorghum beer samples. One ml of each dilution was added to duplicate Petrifilm TM -Aerobic Plate Count (APC) plates. USA) was added to non-Yeast Count media at a concentration of 20mg / ml (Table 1) (Nout et al.. This represented the equivalent of MRS Agar used to determine lactic acid bacteria counts by conventional plating. Pattison et al. To determine the number of aerobic bacteria by Petrifilm TM plating. see text) Petrifilm TM -Yeast and Mold plates .. 2. stored at room temperature at the point of sale and representing the major commercial brands marketed by two large manufacturers. which was found sufficient to inhibit the growth of yeasts in a pilot study (results not shown). according to manufacturer’s instructions. beer samples were serially diluted in peptone supplemented saline containing 20mg / ml cycloheximide. Tenfold serial dilutions were plated in duplicate onto three conventional media by the spread plate technique (Table 1). Table 1 Methodology for microbiological analyses of South African sorghum beer Analysis Conventional plates Atmosphere Growth medium Aerobic plate count (APC) aerobic Lactic acid bacteria count (LABC) anaerobic a Yeast count (YC) aerobic Standard One Nutrient Agar (Biolab) 1 20mg / ml cycloheximide (Sigma) MRS Agar (Biolab) 1 0. found to detrimentally affect the ability to differentiate and accurately enumerate individual colony forming units on the modified-Petrifilm TM APC plates.2% (w / v) potassium sorbate (Unilab) (von Holy et al. 1987).85% NaCl (Biolab. One ml aliquots of an additional tenfold dilution series of beer samples in quadruple strength MRS Broth (Biolab) containing 20mg / ml cycloheximide were added to duplicate Petrifilm TM -APC plates to determine the numbers of lactic acid bacteria. Quadruple strength MRS Broth was therefore made up from its individual components.T. however. U. RSA)] in sterile plastic bags and homogenized in a Colworth 400 stomacher for 2 min (10 21 dilution). The dark colour of commercially available MRS Broth was. and autoclaved. Sample processing Ten ml of each beer sample were added to 90ml peptone supplemented saline [0.7 1 20mg / ml cycloheximide (Sigma) Potato Dextrose Agar (Oxoid). 0. Sterile glucose solution was subsequently added to the other components which resulted in a consistent light-coloured broth facilitating enumeration of colonies and enhancing operator ability to distinguish colonies from beer particles. excluding glucose. 1991) pH 5.1% peptone (Oxoid.

The corresponding counts from Petrifilm TM were 5. MRS Agar and equivalent Petrifilm TM products (Table 1) were selected for characterization. brewer’ s degrees and titratable acidity The mean pH (10 21 dilution) was 3.33 (Novellie and de Schaepdrijver. Titratable acidity (Nout et al. were obtained by conventional plating methods (Table 2).3.0%) of the 419 colonies isolated from Standard One Nutrient Agar. 3.06 to 3. lactate enantiomer configuration (Gawehn and Bergmeyer. 1987) and an indirect measure of acidity using brewer’s degrees [the amount (ml) of 1N NaOH required to neutralize 50 g of beer to a pH of 6. 3. respectively.62 log CFU / ml). Plates showing 30–300 colony forming units (CFU) were counted.53 to 2. 5.05) lower aerobic plate counts (by 0. For Petrifilm TM .68) and the mean titratable acidity was 0.48 log CFU / ml. 1986)] were also determined.118 T.08 (range of 3. 0. By comparison.4. Bacterial populations The majority (369-88. 1987). standard deviations for yeast counts from both plating methods were only 0. PFLABC and PF-YC. 0. Wider ranges and standard deviations in excess of 1 log CFU / ml were obtained for both aerobic plate counts and lactic acid bacteria counts of both plating methods (Table 2).34.96.16 log CFU / ml (Table 2).84 log CFU / ml. 2. lactic acid bacteria counts (LABC) and yeast counts (YC) of 5.5.44 and 7.64% to 0. 0. 2. respectively (Table 2). 1984). gas production from glucose (Schillinger and ¨ Lucke. 6. brewer’ s degrees and titratable acidity The pH of the 10 21 dilution of each beer sample was determined using a Jenway model 3100 pH meter. no statistically significant differences ( p . Statistical analysis Mean bacterial and yeast counts from corresponding conventional and Petrifilm TM media were compared statistically using multifactorial analysis of variance at the 95% confidence level. 2. 1987).1.13 log CFU / ml) than the corresponding conventional plating method (Table 2).85 log CFU / ml for PF-APC.05) lower lactic acid bacteria counts (by 1. pH. 1988) of Standard One Nutrient Agar. The 419 representative isolates thus obtained were purified on the media from which they were isolated and examined microscopically to confirm the absence of any yeast isolates. The bacterial isolates were characterized to genus level according to gram and catalase reactions as well as morphology using the dichotomous key of Fischer et al. The mean lactic acid bacteria counts obtained by conventional plating exceeded the corresponding aerobic plate counts by 0. Conversely. brewer’s degrees were 2. Pattison et al.31 and 7.. The percentage distribution of samples according to bacterial count intervals for both count types and media reflected these trends (Fig. mean aerobic plate counts were. as well as significantly (p . growth at 15 and 458C (Gerber. 2). 3.12). Isolation and characterization of predominant bacterial populations Two colonies from duplicate plates of the highest dilution showing growth (von Holy and Holzapfel.05) were found between yeast counts from the two plating methods and their results corresponded within 0. Colonies from Petrifilm TM plates were removed from the gel medium using a straight inoculation needle.2. Lactic acid bacteria were broadly divided into biogroups based on morphology. almost identical to corresponding lactic acid bacteria counts (Table 2). 1974) and the ability to ¨ hydrolyse arginine (Schillinger and Lucke.3. MRS .79%) with narrow standard deviations (Table 2).72% (range of 0. Microbial numbers Mean aerobic plate counts (APC).01 log CFU / ml. Measurement of pH. Results 3. (1986). / International Journal of Food Microbiology 43 (1998) 115 – 122 All plates were incubated at 258C for 72 h. Petrifilm TM yielded significantly (p . however.60 (range of 2.

0%). Homofermentative lactobacilli were recovered in high and almost equal proportions from Standard One Nutrient Agar (58.3%) and pediococci (7019. 1. 22.16 3.44 A 3. 50% dominance was reduced to 26. Pattison et al. Identities of lactic acid bacteria isolated from the four plating media are shown in Fig. obtained by conventional and Petrifilm TM plating (Table 2) were unexpected in this study.68 0. .70–7. heterofermentative lactic acid bacteria (112-30.79 0. 4. Similar proportions of pediococci were isolated from Petrifilm TM -APC (22.1%) plates.3.8%) plates (Fig. Predominant lactic acid bacteria biogroups present were homofermentative lactobacilli (180-48.9%) and Gram negative bacteria (3-0.30–7.96 1.4%) were lower.27 6.7%). Isolation frequencies of heterofermentative lactic acid bacteria from Standard One Nutrient Agar.1 and 17.3%.85 A 7.64–0. Microbial numbers The high aerobic plate counts and lactic acid bacteria counts of sorghum beer from the South African retail trade.9%. A. where heterofermentative lactic acid bacteria predominated (54.T.07 Standard deviation. 3). This .05) between methods for the same count type.90–6. In support of the expected low bacterial counts prior to heat exchange.53–2. Discussion 4.1%) and modified-Petrifilm TM -APC (53. Means with different superscripts denote statistically significant differences (p . modified-Petrifilm TM -APC and Petrifilm TM -APC plates were 27. 1). however.24 7.40 7. Petrifilm TM -APC and modified-Petrifilm TM APC plates were lactic acid bacteria (LAB). 1).6%) and modified-Petrifilm TM APC (23. pH.25 1.B Agar. brewer’s degrees and titratable acidities of 46 sorghum beer samples Analysis APC / PF-APC(log CFU / ml) LABC / PF-LABC (log CFU / ml) YC / PF-YC (log CFU / ml) pH (units) Brewer’s degrees (ml 1N NaOH) Titratable acidity (% w / w lactic acid) a Petrifilm TM Conventional mean range SD a mean range SD mean range SD mean range SD mean range SD mean range SD A 5.8%). Micrococcus species (12-2. Tlhoaele (1996) reported APC and LABC of .1. 1989).4%).00 log CFU / ml of wort after the straining stage (Fig.79 1.02 2.04–7.0%). Fig.06–3.60 2.48–8. respectively.31 B 1.34 B 1.39 1. Petrifilm TM -APC (57.84 A 7.16 5. Bacteria introduced into the mixture during the first and second souring stages are presumably destroyed during the ’cooking’ and ’heating’ stages which serve to gelatinize starches and destroy vegetative forms of bacteria (Fig.72 0. 3).40–8. Subsequent stages prior to cooling of the wort in the heat exchanger are performed at 608C to take advantage of the protective action of the temperature in preventing possible microbial contamination of the wort (Haggblade and Holzapfel. The remaining isolates comprised Bacillus species (358. corresponding pro- portions from Standard One Nutrient Agar (11. 3.08 5.08 3.06 0. 0.12 0.96 1.08 0.19 0. / International Journal of Food Microbiology 43 (1998) 115 – 122 119 Table 2 Mean microbiological counts.6% on conventional MRS Agar.

the analysis of beer samples of different ages would also explain the high variation. 4. increased significantly to between 5 and 6 log CFU / ml after passing through the heat exchanger. which is added to the wort to initiate alcoholic fermentation (Fig. exhibited APC and LABC of between 4 and 5 log CFU / g (Tlhoaele. the high yeast counts and corresponding low standard deviations obtained in the present study were expected since Saccharomyces cerevisiae is inoculated into the wort in high numbers. 1). Previous studies on comparative evaluations between Petrifilm TM and conventional methods to . 2. 3. indicating insufficient cleaning and sanitizing of this piece of equipment on the production line. 1).0 and 208C. MRS Agar (B). shortly before the beer is distributed to the retail trade. at a pH and temperature of ca. this could not be determined conclusively since this information was not available from the retailers. to initiate alcoholic fermentation. Petrifilm TM -APC (C) and modified-Petrifilm TM -APC (D). respectively (Fig. Pattison et al. Fig. thus providing an additional source of bacterial contamination of the final product. In addition. however.120 T. Percentage distribution of 46 retail sorghum beer samples by conventional (A) plating and Petrifilm TM (B) methods according to bacterial counts. The high bacterial counts with corresponding wide ranges and large standard deviations obtained in the present study (Table 2). the yeast starter culture is pitched into the wort at optimum conditions for yeast growth. The bacterial counts of the wort. 1996). In the same study. dry brewer’s yeast. Furthermore. / International Journal of Food Microbiology 43 (1998) 115 – 122 Fig. however. indicated that contamination of the sorghum beer was incidental and the extent to which it occurred was variable. Conversely. Identities of 369 predominant lactic acid bacteria from Standard One Nutrient Agar (A).

2. the significantly higher count (by 1. indicated the presence of considerable proportions of potential spoilage bacteria in the product tested.. 4. Conclusion High bacterial counts of sorghum beer at the point of sale indicated high levels of contamination of the product before distribution to the retail trade and possible growth of these populations in the product. 3. 1985. The low pH of the beer reportedly inhibits or kills pathogenic and most anaerobic endospore-forming bacteria.13 log CFU / ml) obtained on MRS Agar compared to that obtained on the modified-Petrifilm TM -APC (Table 2). Based on these results.3–0. It is also likely that the concentration of glucose in MRS Agar is higher than that present in the modified-Petrifilm TM -APC plates. 0. In the present study. the manufacturers’ equipment and the factory environment. 4. .6% in sorghum beer (Novellie and de Schaepdrijver. 1986).0). The source of these bacteria was not determined in this study and would require microbiological surveys of product line samples. However. Smith et al. Chain and Fung. This provided further evidence that the modification to the Petrifilm TM -APC with quadruple strength MRS Broth was insufficient to mimic the MRS Agar. 1989). Bailey and Cox. Beuchat et al. (1997).. Bacterial populations The high proportions of LAB in the sorghum beer indicated the ability of these bacteria to survive and grow in an acidic environment. as reported by Linton et al. heterofermentative lactic acid bacteria and pediococci in the sorghum beer at the point of sale. / International Journal of Food Microbiology 43 (1998) 115 – 122 determine the microbiological quality of different food commodities showed high correlations between the two methods (Ginn et al.08 for the beer samples tested in this study was lower than that reported by Haggblade and Holzapfel (1989) (pH 4. Mean titratable acidity in this study exceeded previously reported levels of 0. Pattison et al. Petrifilm TM -APC and modified-Petrifilm TM -APC plates were similar.T. thus also accounting for the differences shown in Fig. the modified-Petrifilm TM -APC resembled the Standard One Nutrient Agar and Petrifilm TM -APC. showed that bacterial counts with Petrifilm TM were significantly (p . would probably have resulted in different predominant populations being present on the plates of highest dilution showing growth. This is probably due to the observed presence of high proportions of potential spoilage bacteria in the product tested. The possibility of pasteurizing sorghum beer after alcoholic fermen- 121 tation in order to eliminate spoilage populations and thus extend the shelf life of the product may be considered. The presence of homofermentative lactobacilli. pH.3. concentrations as high as 1% lactic acid can reportedly be tolerated before the beer is rendered unacceptable to consumers (Haggblade and Holzapfel. Furthermore. 1984. 1987. certain mesophilic bacteria may be better resuscitated in conventional media due to a more optimal water activity and oxidation / reduction potential. Proportions of LAB biogroups isolated from Standard One Nutrient Agar. the increased salt concentration introduced to the Petrifilm TM via the diluent may have resulted in the lower bacterial counts.. Thus in terms of populations recovered. 1989). thus ensuring the safety of the consumer (Haggblade and Holzapfel.05) lower than counts obtained by conventional means (Table 2). The difference in counts was particularly pronounced for the lactic acid bacteria counts (Table 2). 5. For example. Results of the present study. which produce additional acid and thus result in the pH drop. Furthermore. the PF-APC and PFLABC were found to be almost identical (Table 2). The lactic acid concentration of the beer by measurement of brewer’s degrees was in agreement with figures reported by Novellie and de Schaepdrijver (1986). brewer’ s degrees and titratable acidity The pH value of 3. It is speculated that differences in the composition of the conventional media and corresponding Petrifilm TM may have resulted in the under-recovery of bacteria with Petrifilm TM . 3). however. we speculate that the modification of the Petrifilm TM -APC with the addition of quadruple strength MRS Broth was insufficient to mimic the conventional MRS Agar in determining the number of lactic acid bacteria of sorghum beer. provided that this treatment would not unacceptably change the character of the beer (Haggblade and Holzapfel. 1991. but differed from those isolated from MRS Agar (Fig. 1989). 1998).

Growth of Bacillus cereus in soyabean tempeh. Townsend. R.) Microbiology of Fermented Foods.. and Conventional pour-plate methods for enumerating aerobic microorganisms in foods. 50. pp. Tlhoaele. J. R. Danisavich. ¨ Schillinger.K. J. 14–18. Food Microbiol. B.R.S. 199–208. J. 4.. J. However.B. Odunfa.H.R. 12. Afr. 1991. Meissner. P. Novello. 1987. 7 (2). M... S. Fung. Weinheim.C.W. J. Holzapfel. 1996.S. The influence of extrinsic factors on the microbiological spoilage pattern of ground beef. N.J.L. Chain. 28. Comparison of Redigel. Afr. L..U. 6. 47. Curiale. for financial and logistical support. As an accelerated alternative to conventional methods. P.. 293–301. Bacterial populations associated with the production of sorghum beer in a commercial brewery. Comparison of conventional plating methods and Petrifilm for the recovery of microorganisms in a ground beef facility. Fox.E. J. Jooste. 1044–1045.P.. van der Walt. W. acid fermented vegetable products and fermented meat products. M.. Dairy Sci. 18. P.R. T.H..F. Afr. Food Protect. S. J.. Muriana.. de Schaepdrijver. 1984.. D. 643–644. B. King. statistically significant differences ( p . Kaffircorn malting and brewing studies II-Studies on the microbiology of Kaffir beer. J. Evaluation of the 3M dry medium culture plate (Petrifilm TM SM) method for determining numbers of bacteria in raw milk.. 1986. 1974.H. FEMS Microbiol. F.A. 1991. Industrialization of Africa’s indigenous beer brewing. M.Sc. Food Protect. H. A. Amsterdam. Elsevier Applied Science Publishers. Industrialization of Indigenous Fermented Foods.... Bonants-van Laarhoven.. 387–390.05) in bacteria counts between Petrifilm TM and conventional enumeration methods. J. (Ed.G. pp. . 269–280. 1987. Methoden der Enzymatischen Analyse II. 1084–1088. J.O. Holzapfel. Lactic-acid fermentation as a low-cost means of food preservation in tropical countries. Bergmeyer. U.. Fischer. 1986. 155–191.. In: Ber- gmeyer. Effects of pasteurization and storage temperature on vacuum-packaged vienna sausage shelf-life. Fox. Elsevier Science. 105–113..R. 1963. 1987. Evaluation of the Petrifilm SM and VRB dry media culture plates for determining microbial quality of poultry. 1987. 1985. V. South Africa. Busta.. T. F. L. W. Verlag Chemie.. New York. as well as differences in composition of lactic acid bacterial populations were encountered in this study. 61. L. Petrifilm TM . 191–283. Food Protect.H. M.A. Gawehn. 60.A. Int. 46.B.F.. von Holy.C. This observed under-recovery of lactic acid bacteria should be taken into consideration when utilizing Petrifilm TM for routine enumerations of products containing predominantly lactic acid bacteria..J.. Pretoria. In: Woods. 87. (Ed.. Food. Comparison of the SimPlate TM Total Plate Count method with Petrifilm TM . pp. South Africa. 1984.).L.L. K. 28. In: Adams.. P.. Food Protect. R..... 4. Packard. T. Holzapfel. P. 0. Sci. Food Protect. Bantu beer-food or beverage? Food Ind.H. University of Pretoria.D. Likin. 369–379. Linton. H. 208–211. In: Steinkraus K.122 T. Modern developments in traditional African beers. Reilly. H. B. J. Identification of lactobacilli from meat and meat products. J. Agric. 73–157. 753–759. T.. (Ed. S. A. Novellie. D... Cooke. R. 54. 1966. Lucke. Lawlis. Microbiology Department.. 1988.. Gangar. T. Cox. / International Journal of Food Microbiology 43 (1998) 115 – 122 Characterization of the predominant LAB isolated from plates of the beer samples indicated the presence of potential spoilage bacteria. V. W. Food Microbiol.. 1997. Acknowledgements The authors are indebted to 3M (South Africa) (Pty) Ltd. S. L. Twiddy. Int.. Brewer Distiller 1. Novellie.. cultured dairy products. University of the Witwatersrand. Spiral Plate System Isogrid and Aerobic Plate Count Agar for determining the numbers of aerobic bacteria in selected foods. Okwusoa. Haggblade.).). J. S. 1998. Rev. Int. Food Microbiol.M.. V..J..U. J.L.. Food Protect. African fermented foods.. Taksonomie van melksuurbakteriee¨ geassosieer met vakuumverpakte vleisprodukte. W. G. Eisel. 23. 1989. Bantu beer: Popular drink in South Africa. Pattison et al. Evaluation of rapid plate counting techniques for enumeration of psychrotrophic bacteria in raw milk..Sc Honours Dissertation. 48. Marcel / Dekker. Progress in Industrial Microbiology. Copeland..M.. D. Smith.J. Johannesburg. Petrifilm TM offers the advantages of time and cost efficiency.S. D-(2) Lactat.Y. Redigel TM . Ginn. 33. Beuchat. Comparison of a dry medium culture plate (Petrifilm TM SM plates) method to the aerobic plate count method for enumeration of mesophilic aerobic colony-forming units in fresh ground beef. L. von Holy. F. Smith.S. 137–141. Gerber. Sci. 1985. References Bailey. Department of Microbiology and Plant Pathology. E. (Agric) Thesis. Beernink. Nout...S. S.. Novellie. (Ed. pp.G. D.. for example sorghum beer. Amsterdam. 1956. C. 2. 1538–1541.E.