Population Dynamics of Constitutive Microbiota in BAT Type Fermented Milk Products

Transcription

1 70 Journal of Food Protection, Vot 58, No.1, Pages Copyrighl, International Association of Milk, Food and Environmental Sanitarians Population Dynamics of Constitutive Microbiota in BAT Type Fermented Milk Products L. M. MEDINA and R. JORDANO* Department of Food Science and Technology, University of Cordoba, E Cordoba, Spain (Received February 28, 1994/Accepted June 7, 1994) ABSTRACT Seventy samples from two batches of commercially produced fermented milk were analyzed for constitutive microbiota. Five samples were tested on the day of collection and the others analyzed after 10, 17, 24, 28, 31 and 36 days of storage at 7 C. M]7 agar was used to enumerate Streptococcus salivarius subsp. thennophilus. MRS (de Man, Rogosa, Sharpe) agar was used to enumerate Lactobacillus acidophilus. For the enumeration of Bifidobacterium spp. MRS agar to which antibiotic substances were added was used. In the case of Batch A, streptococci, lactobacilli and bifidobacteria were initially present at levels of x 10 6,0.677 X 10 6 and x 10 6, respectively. The initial ph of 4.48 decreased to 4.12 after 24 days. For Batch B, streptococci, lactobacilli and bifidobacteria were initially present at levels of 858 x 10 6, X 10 6 and x 10", respectively. The initial ph of 4.23 decreased to 4.07 and 3.84 after 24 and 36 days, respectively. The fall in the counts throughout refrigerated storage time showed some irregularities which, in our opinion, indicate a lack of control over the ecological aspects which affect the interrelation between the microorganisms involved. The statistical significance of the differences between the different trials was established throughout the time. Key Words: BAT-type products, fermented milks, bifidobacteria, Lactobacillus acidophilus The large number of fermented milk products continuously appearing on the international market justifies the routine task of their definition, characterization and testing, especially when it is realized that in many countries, among them Spain, legislation is a step behind reality and only specifications concerning yogurt are known. Kurmann, et al (10) have examined a great number of fermented fresh milk products, including aspects of their microbiology and technology. This publication seems to us to be of great value for a definition of the specific features of each product. To this effect, we have begun to analyze the growth of the constitutive microbiota in some fermented milk products by using selected strains of intestinal bacteria. We have two basic aims: a. To check the possible control exercised by manufacturers on the overall behavior of the microorganisms involved, by means of a study of microbiological growth over a period of time and under refrigeration. b. Because of this growth, and of parallel tests in which hygiene quality, risks of contamination, and properties derived from the presence of beneficial microorganisms as integral parts of the constitutive microbiota could be evaluated, in future work criteria would be established for defining an ideal shelf life for each case, not simply applying that corresponding to yogurt, as happens in many countries. In a previous report (12), we studied a product whose constitutive microbiota was made up of yogurt microorganisms in combination with Bifidobacterium spp. On this occasion the objective was the study of the dynamics of population of constitutive microbiota during refrigerated storage in two BAT (Bifidus-Acidophilus- Thermophilus) type products called Bifidus-Active, produced in Spain. MATERIAL AND METHODS Preparation and sampling. Seventy samples from two batches of commercially-produced fermented milk were transported to the laboratory in iceboxes, maintained at 7 C, and analyzed for constitutive microbiota. Five samples were tested on the day of collection and the others were analyzed after 10, 17, 24, 28, 31 and 36 days of storage at 7 C. Samples were aseptically removed from containers and diluted by mixing 10 g with 90 ml of 0.1 % bacto-peptone (Difco Laboratories, Inc., Detroit, MI). Further dilutions were made as required. Enumeration of S. Streptococcus salivarius subsp. thermophilus. M]7 agar (26) was used to enumerate S. salivarius subsp. thermophilus. The ph of the medium was 7.0 ± 0.1. The inoculated plates were incubated at 37 C for 48 h. Under these conditions S. salivarius subsp. thermophilus forms lenticular colonies with a diameter of 1-2 mm. The counts were expressed as CFU/g. Enumeration of Lactobacillus acidophilus. MRS (de Man, Rogosa, Sharpe) agar (Difco) acidified to ph 5.4 at 25 C was used to enumerate L. acidophilus. The plates were incubated at 37 C for 3 days in an anaerobic culture jar (H/C0 2 ). The counts were expressed as CFU/g. The lactobacillus colonies

2 DYNAMICS OF POPULATION IN BAT-type PRODUCTS 71 identified were corroborated by observation under a microscope. Lactobacillus acidophilus were gram-positive rods with rounded ends (/). Enumeration of Bifidobacterium spp. MRS agar (ph 6.23) heated to 50 C was used although in the case of Batch A, 1 ml of a solution (2 ~g/ml) of Nafcillin (Sigma Chemical Co., St. Louis, MO) was added to each plate, while in Batch B, a 5% solution of Dicloxacillin (Sigma) had been previously added to the MRS agar at the moment of its use at a proportion of 1%. Both antibiotic solutions were previously sterilized by filtering through Millipore filters with pore diameter of 0.45 ~m. Incubation was in anaerobic jars with H/C0 2 atmosphere (Oxoid), at 37 C for 3 to 5 days. The bifidobacteria colonies identified were lenticular, or irregularly shaped, and were corroborated by observation under a microscope. Both in the case of the bifidobacteria and in that of L. acidophilus the supposed identification that we had made was validated by the laboratory of the Spanish Collection of Type Cultures (CECT), Department of Microbiology, University of Valencia, Spain. The results were expressed as CFU/g. ph determination. The ph was measured using a digital ph meter (Crison, Model 501) with a combination electrode. TItratable acidity estimation. Acidity was estimated by titrating 9 g of sample diluted with 18 ml of CO 2 -free water with 0.1 N sodium hydroxide (NaOH) according to Standard Methods for the Examination of Dairy Products (/1). The results were expressed as lactic acid percent. Statistical analysis. Data were analyzed using the general linear models procedure of Statistical Analysis Systems (SAS) (20). Differences were considered to be significant at P <.05. RESULTS AND DISCUSSION Counts of constitutive microbiota in Batches A and B are expressed in Tables 1 and 2, respectively. Streptococcus salivarius subsp. thermophilus. In both batches, S. salivarius subsp. thermophilus was the most numerous component. In Batch A, the initial count was x 10 6 CFU/g; at 10 days the numbers only fell by 2.08% and the maximum was in the range of 257 x 10 6 CFU/g. At the end of the shelf-life there were still x 10 6 CFU/g. In all this time, the most notable fall was of only 7.26%, between 10 and 17 days. The most important fall was noted between the 28-days trial and the one made at 36 days, which showed a fall of 56.58%. The counts were analyzed statistically by single-factor variance analysis and the Tukey test (20), with a significant fall in the values, although this was not very noticeable until the fourth week of refrigerated storage. It should be borne in mind that, in the order of the trials themselves, the results of the analysis for the time comprising the entry on the market of the product and 24 days later did not show any statistically significant differences when considering this group of values. However, and without distinction, we could refer to another group, not statistically different, which included the numbers corresponding to 17, 24 and 31 days. If we observe that there does exist a significant difference between the values corresponding to the first two trials and that of 31 days (or that of 28 days), we shall be completing a progression that goes slightly downwards, as we have already mentioned. For B~tch B, the initial trial gave 858 x 10 6 CFU/g with a maximum of 10 9 CFU/g. At 10 days the count is somewhat higher (8.97%). The values determined in successive trials began decreasing during the rest of the product's shelf-life, exhibiting 714 x 10 6 CFU/g on day 24 (end of the shelf-life) which was 83.2 I % of the initial value. Subsequently, and up to 36 days, the numbers kept going down, with the most notable reduction being at between 31 and 36 days of refrigerated storage (26.66%). Because of the Tukey test, the differences between different trial groups turned out to be statistically significant. It was not surprising that there was an increase in the second analysis with respect to the first, taking into account the growth of other microorganisms studied in other products. Even so, the specific differences between the first two counts were not statistically significant, which does not seem to us to be symptomatic of any anomaly. However, the fact that the first count was higher justified this value being statistically nondifferentiable from the one of successive trials (up to 24 days) which, in tum, are indeed statistically TABLE 1. Number and percentage of surviving constitutive microbiota in BAT type fermented milk products (n = 35) during refrigerated storage (J0C). Batch A. Storage time ph TA' S. salivarius subsp. L. acidophilus Bifidobacterium spp. (days) thermophilus a a a 100-' a a " a b b a a. b ' b ' ' b lob., ' b d ' O b 3.85 ITitratable acidity expressed as percent lactic acid. 2Mean values x 10 6 expressed as CFU/g from 5 samples. 3Percentage of surviving microorganisms. 4Maximum shelf-life. a.b,c.dmeanswith the same superscript in the same column do not differ (P<.05)

3 72 MEDINA AND JORDANO TABLE 2. Number and percentage of surviving constitutive microbiota in BAT type fermented milk products (n = 35) during refrigerated storage (l C). Batch B. Storage time ph TAl S. salivarius subsp. L. acidophilus Bifidobacterium spp. (days) Ihermophilus a,b IO(}' " " ' ' b b,c b c b,c b d 7, c G b d c,d b O.077 d d b d 8.82 'Titratable acidity expressed as percent lactic acid. 2Mean values X 10 6 expressed as CFU/g from 5 samples. 'Percentage of surviving microorganisms. 4Maximum shelf-life. a,b,c,dmeanswith the same superscripts in the same column do not differ (P<.05) different (and lower) from the maximum determined at 10 the incubation period of the inoculate was foreseen (16). In days of refrigerated storage. If we exclude the result from BAT type products, its growth for 14 days under refrigerathe first analysis, we could take into consideration a third tion at ph 4.5 has been described: only at 11 days does its group of values which were not statistically different, ex- number begin to slightly decline from a range of hundreds tending to the one corresponding to day 31. of millions to tens of millions (21), so that we can consider In both products, a highly significant correlation was the products studied, as far as the streptococci are confound between the counts of S. salivarius subsp. cemed, as being within expectations. They may even have thermophilus and the acidity. With regard to the ph, it was greater numbers since the ph was continually lower than the same except that the correlation in Batch A was only the levels studied by the cited author. significant at 95%. In a previous paper (12), the initial population of streptococci, X 10 6 CFU/g, except in the test carried out at 10 days, underwent a moderate fall during the storage period. After 51 days the number of viable cells still represented 69.5% of the initial number. It was in the last determination that a very considerable reduction was noticed since between day 51 and day 84 the population of strep~ tococci decreased by 58.7%. In an experiment on Biogarde (a fermented milk product) without fruit, in approximately a little over 80% of cases, S. salivarius subsp. thermophilus was to be found in the range of loscfu/ml (23). In commercial yogurt kept for 40 days at 7 C, S. sa/ivarius subsp. thermophilus and Lactobacillus delbrueckii subsp. bulgaricus underwent a statistically significant numerical fall and' these microbiota, after 30 days, were reduced by close to 50% (6). The population of viable organisms in yogurt increased initially after manufacture, rea~hed ~ maximum number, and then decreased during their refngerated storage (4). Likewise, in one of the three cases studied, the possibility that the decrease in the number of lactobacilli was faster than 'that of the streptococci was also confirmed. Manufacturing practices and the variety of starting cultures used by the different manufacturers had the greatest effect on the survival of S. salivarius subsp. thermophilus. The presence of S. salivarius subsp. thermophilus is esteemed by us to be important in products such as these, especially as a consumer of oxygen, which creates a much more appropriate atmosphere for the other components of the constitutive microbiota. That it is the most numerous component causes no surprise, and the trend described in Lactobacillus acidophilus An investigation on L. acidophilus revealed appreciable differences between the batches analyzed (Tables 1 and 2). In Batch A we started with a lactobacillus count of X 10 3 CFU/g which was exceeded by that obtained at 10 days, reaching 780 X 10 3 CFU/g with a maximum in the range of 940 X 10 3 CFU/g. From this moment on, the numbers fell sharply (between 10 and 17 days by 73.67%, and between 17 days and the end of the shelf-life by 91.39%). The counts went on falling in successive trials except in the one carried out at 36 days, which practically maintained the values established in the trial immediately prior to it. A single-factor variance analysis and the Tukey test indicated this decrease showed statistically significant differences between the first two analyses and the one carried out at 17 days and, in its tum, between the latter evaluation and subsequent analyses. The initial count corresponding to Batch B was X 10 6 CFU/g was slightly lower than that recorded at 10 days of refrigerated storage, 200 X 10 6 CFU/g with a maximum of 256 X 10 6 CFU/g. From the third trial (17 days) until day 28, the counts underwent a slight recovery, although they were clearly lower than those recorded in the first two trials. The last determinations maintained a downward trend, notwithstanding some fluctuations. A statistical study corroborated the hypothesis that had been proposed as was demonstrated by the significant differences between the results of the first two trials and the remainder. The downward progression of the results was not exempt from irregularities in the form of tiny fluctuations, although those values were not statistically different from each other (P<.05). JOURNAL OF FOOD PROTECT/ON, VOL. 58, JANUARY 1995

4 Bifidobacterium spp. In Batch A, at the end of the shelf-life a count of bifidobacteria of x 106 CFU/g was recorded, which signified 36.84% of the initial total and a fall of 62.82% with respect to the previous value. There was a downward trend up to 36 days (Table 1). A single-factor variance analysis and the Tukey test again showed significant differences between the values determined in different trials. To be specific, there were no DYNAMICS OF POPULATION IN BAT-type PRODUCTS 73 Likewise, in the two products studied, a highly signifi- significant differences in the counts corresponding to 0, 10 cant correlation coefficient (P<.OO1) was established be- and 17 days, but there were differences between these and tween L. acidophilus and the ph, and also between that the remainder in an obviously downward direction. microorganism and the acidity. The popularity of yogurt Concerning Batch B, we started with a count of suggests its possible utilization as a carrier of other acid x 106 CFU/g, similar to that of the previous product. lactic bacteria (18). However, L. acidophilus had difficul- The fall in counts was effective up to 36 days of refrigerties in surviving (2). ated storage, the last day accounted for in the experiment The reason that these lactic acid bacteria are affected (Table 2). may be due to acidity; in diverse products comparable to Once again, as in all the previous cases, there were acidophilus milk the number of L. acidophilus cells fell to significant differences between the different trials. Although levels below the desirable ones in a matter of days (24), for the whole of the experiment growth was irregular, within reflecting extreme sensitivity to lactic acid levels above what is specifically the commercial shelf-life there was a 0.6%, which emphasizes the need for strict control during more notable fall of bifidobacteria than of the other microthe manufacturing process. Another cause might be the fact organisms, given that the values obtained in each trial, even that some strains of L. delbrueckii subsp. bulgaricus, an up to the expiration date, were significantly different and habitual starting strain not only in yogurt but also in other each one lower than the one before (Table 2). fermented milks, may produce levels of hydrogen peroxide In both batches there was a highly significant correlawhich inhibit certain bacteria (25). tion (P<.OOl) between the counts corresponding to Lactobacillus acidophilus inoculated in pasteurized milk bifidobacteria, to the ph and to the acidity, in each case. remained viable for 2-3 weeks at a low temperature (23). Before carrying out this present study, we made an evalu- Also, in ice cream, 1.5 X 10 8 CFU/g were counted immedi- ation of the survival of constitutive microbiota in a comately after freezing (5). An account of the survival of L. mercial fermented milk containing bifidobacteria, the first acidophilus in acidophile yogurt was published and for the count of which was 7.4 x 106 CFU/g, afterwards continuing natural variety gave values of 95 x 10 5, 76 X 10 5 and 40 x 10 5 in a downward direction (12). CFU/g in an initial trial and after 7 and 14 days under With regard to the products which are even closer to refrigerated storage, respectively (18). There were some es- those being studied, the presence of bifidobacteria in sential differences regarding the counts which were fore- Biogarde was evaluated in random ranges and in only seeable for the Cultura, such as 2-4 x 10 8 for L. acidophilus slightly over 40% of the cases were there counts of between and 1-2 X 10 8 CFU/ml for Bifidobacterium bifidum. 106 and 10 8 CFU/ml to be found (23). These data do not The presence of L. acidophilus in an experiment gave contradict those of Klupsch (7), who situated at 10 8 /ml the values within or higher than the range of 10 8 CFU/g in over level of viable cells of L. acidophilus and at 106/ml that of 90% of the cases, and a little over 70 % were in the range B. bifidum. Also, Glaser emphasizes the importance of of 10 9 CFU/g (23). In our experiment, only product B was maintaining a count of at least 10 7 CFU/g of characteristic found to be over 10 8 CFU/g during the first two counts. microbiota in other fermented milks different from yogurt; In any investigation of L. acidophilus in fermented milk in the case of bifidobacteria, this count should be at least products which include it, not only should its presence be 106 CFU/g (3). The number of bifidobacteria in fermented confirmed, the strain used noted, and a quantification made, milks, BAT type and with a ph of between 4.3 and 4.7, but also this analysis should be contrasted with the aims of may decrease by 2 log cycles during 1-2 weeks of storage the product or its adjustment to the minimum levels re- (16). An adequate count for fermented milk would be 106_ quired in order to be able to promote certain products as 10 8 bifidobacteria per milliliter (22). probiotic or as being beneficial to dietetics and to health. For the preparation of bifidus milk powder, the growth We consider that in the case of the products under study of Bifidobacterium longum in a nitrogen atmosphere turned there was an obvious first and important decrease in the out to be practically independent of the oxygen previously lactobacillus counts in the period of between 10 and 17 dissolved in the medium (14). The four strains studied dedays from the beginning of the experiment. In Batch A, the creased in numbers after coming out of certain levels of "therapeutic minimum" advised by Schuler-Malyoth (21) ph, with strain type E194b being the most stable one. Counts was only exceeded up to the analysis made at 17 days in were also made in other products such as ice cream to refrigerated storage, while in the case of Batch B the con- which health-benefiting microorganisms were added, and tribution of these microorganisms was positive with a view 2.5 x 10 8 CFU of bifidobacteria were detected per gram of to covering the probiotic aspect of the food. ice cream after freezing (5). The viability of bifidobacteria in dairy products is still an important problem to be solved. With adequate acidity, free from the influence of oxygen and at refrigeration temperatures, there should be no problems for them to remain viable even at 4 weeks (17). There have been experiments done in which the various strains of bifidobacteria were tested in solutions at various ph values and kept under refrigeration. Differences were found between them, although most notable was the demonstration that the decrease in the counts was less gradual as the solution has a lower ph (13).

5 74 MEDINA AND JORDANO In acidified yogurt-like products, the viability of the bifidobacteria is reduced to a considerable degree, depending on the ph. Values of ph of 4.2 and 4.5 have been regarded as being critical for the survival of B. longum (21,9). Reuter (17) confirmed these data, considering that bifidobacteria counts decreased in fermented milks at around I logarithmic cycle per day if the ph fell below 4.4. However, this author pointed out that in a Japanese product of the yogurt-like type adjusted to a ph of 4.25, the survival of bifidobacteria has been demonstrated, with a reduction in the counts of only I log cycle in 10 days. In the same way as we proposed for the lactobacilli, we consider that in any investigation of bifidobacteria in products with bifidus in general and in BAT type products in particular, not only should the presence of organisms of this genus be confirmed, but also the specific species utilized should be identified. Also, an adaptation of the counts to some objectives or minimums should be estimated which could be legally established so that there would be a reliable response to the promotion and sale of these preparations as contributions of beneficial properties in the fields of dietetics or therapeutics. For instance, it is obvious that in the products studied by us, none in either batch reached the expiration date with at least 10 6 bifidobacteria per gram, something which the dairy industry should try to achieve, although they did go beyond the 24 days of shelf-life, exceeding the therapeutic minimum of Schuler-Malyoth (21). The bifidobacteria growth was irregular, especially when the experiment was prolonged (over I month under refrigerated conditions). Interrelation between microorganisms. By means of an analysis of single variance, the two batches were differentiated, and the differences between them have been catalogued according to the levels of each microorganism (Table 3). Regarding the introduction of L. acidophilus and B. bifidum in other dairy products, it is considered that the changes of ph during storage in its commercial shelf-life were not significant (8). However, there are no reliable data concerning the survival of these microorganisms in the products resulting from that fermentation (18). The only real obstacle to this utilization was their sensitivity to acidity. This is of great importance for the survival of the microorganisms during the storage and distribution of these products. Lactobacillus acidophilus, for instance, ought to be at levels of viability of at least 10 5 CFU/ g in the product for it to have any therapeutic relevance (15). One solution would be to maintain these products at a range of low acidity or to inoculate them with a higher number of microorganisms (18). TABLE 3. Comparison between batches A and B of the constitutive microbiota. Microorganisms S. salivarius subsp. thermophilus L acidophilus Bijidobacterium spp. Level of Significance P<.OOI P<.OOI Not significant In a preliminary study (12), the initial counts of streptococci, lactobacilli (L. delbrueckii subsp. bulgaricus) and bifidobacteria were x 10 6, 50.7 X 10 6 and 7.4 X 10 6 CFU/g respectively. The level of streptococci we found in Batch A was similar, but that of the lactobacilli and bifidobacteria was clearly higher in both batches. After 24 days of refrigerated storage, S. salivarius subsp. thermophilus was still maintained 89.3% of the initial figures, while the Bifidobacterium spp. were reduced to almost half and the lactobacilli were reduced to 14.6% of the initial numbers. At the end of the experiment, the viable streptococci were 2.2 x 106 CFU/g (0.8%) and the rest of the microorganisms studied were very sparse. The initial ph was 4.57 and fell to 4.25 and 3.81 after 24 and 84 days. Schuler-Malyoth (21) described a decrease in the number of viable bacteria in the constitutive microorganisms in a BAT type product. Lactobacillus acidophilus and B. bifidum started with similar numbers within a range of around 10 8 cells/ml and it was seen that the numbers declined as from 4-6 days of refrigerated storage when kept at a ph of 4.5. Bifidobacterium bifidum counts were reduced faster than those of the lactobacilli and at 10 days were already below the therapeutic minimum recommended by this author. Lactobacillus acidophilus fell below this level at 14 days. A general recommendation would be that for this type of product, with culture mixtures with therapeutic implications, inoculates be made in situ on reconstituted powdered milk (19). The fact that none of the species examined was especially competitive in the presence of other microorganisms underlined the importance of the manufacturing conditions being exceptionally pure, and the initiating milk product should be free from any vegetative cells of other bacteria. For these authors, a high rate of inoculation is an assurance that the final product will contain an abundant number of viable cells (l /ml of B. bifidum), but this level is also produced in a short period of fermentation; the ph of the subsequent product is normally about 4.6 and, therefore, incubation could terminate at This comparatively high ph could be essential to assure the survival of the cultures selected throughout the whole commercial shelf-life of the product, as these authors insist on the extreme sensitivity of these microorganisms to acidity. For that reason attention to the isolation of B. bifidum strains tolerant to ph values of under 4.0 becomes important (16). Taking into account the consideration of the small alterations of the ph during refrigerated storage in which the production of lactic acid was severely restricted (8), when the final product began with a ph of 4.6 after the cooling and packaging stages, the constitutive microbiota remained viable throughout a possible shelf-life (19). Finally, the numerical relation which existed between the microorganisms participating in this type of product could be variable (16) and, in the specific case of the products studied in our experiment, they turned out to be irregular. REFERENCES I. Calicchia. M. L., C. I. Wang, T. Nomura, F. Yotsuzuka and D. W. Osato Selective enumeration of Bifidobacterium bifidum, Enterococcus faecium and streptomycin-resistant Lactobacillus acidophi/us from a mixed probiotic product. J. Food Prot. 56:

6 DYNAMICS OF POPULATION IN BAT-type PRODUCTS Gilliland, S. E. and M. L. Speck Instability of Lactobacillus acidophilus in yogurt. J. Dairy Sci. 9: Glaser, H Living bacteria in yogurt and other fermented milk products. Eur. Dairy Magazine 5: Hamann, W. I. T. and E. H. Marth Survival of Streptococcus thermophilus and Lactobacillus bu/garicus in commercial and experimental yogurts. J. Food Prot. 47: Hekmat, S. and D. McMahon Survival of Lactobacillus acidophi/us and Bijidobacterium bijidum in ice cream for use as a probiotic food. J. Dairy Sci. 75: Jordano, R Evoluci6n de la microflora esencial del yogur comercial para su conservaci6n. Arch. Zootec. 33: Klupsch, H. J Sauermilcherzeugnisse. Th. Mann GmbH, Hildesheim. 8. Klupsch, H Bioghurt-Biogarde; Acidified milk products with optimum qualities. North Eur. Dairy 1. 2: Kurmann,1.A Die Bifidobakterienkulturim milchverarbeitenden Betrieb. Deutsche Molkereizeitung 194: Kurmann, J. A., L. J. Rasic and M. Kroger Encyclopedia of fermented fresh milk products. AVI Press (Van Nostrand Reinhold), New York, NY. 11. Marshall, R. T. (ed.) Standard Methods for the Examination of Dairy Products. 16th ed. American Public Health Association, Washington, D.C. 12. Medina, L. M. and R. Jordano Survival of constitutive microflora in one commercial fermented milk containing bifidobacteria during refrigerated storage. J. Food Prot. 56: Mutai, M., M. Mada, K. Nakajima, S. Takahashi, T. Nakao, K. Shimada, T. Iijima and Kunitachi Lebensfahige Bifidobakterien enthaltendes, fermentiertes Milchprodukt und Verfahren zu seiner Herstellung. German Federal Republic Pat. 2,656, Nagawa, M., J. Nakabayashi and S. Fujino Preparation of the bifidus milk powder. 1. Dairy Sci. 71: Nahaisi, M. H. and R. K. Robinson Dairy Ind. Int. 50:16. Cited by R. K. Robinson. (1987). 16. Rasic, L. J. and J. A. Kurmann Bifidobacteria and Their Role. Bilkhliuser Verlag, Basel, Switzerland. 17. Reuter, G Bijidobacterium characteristics and classification. pp In John Libbey, (ed.). Fermented Milks: Current Research. International Congress, Paris. 18. Robinson, R. K Survival of Lactobacillus acidophilus in fermented products. S. Afr. 1. Dairy Sci. 19: Robinson, R.K. and A. Y. Tamime Microbiology of fermented milks. pp In. R. K. Robinson, (ed.). Dairy Microbiology. Elsevier Sci. Pub., Ltd., London. 20. SAS User's Guide: Statistics, Version 6.0 Edition Statistical Analysis System Institute, Inc., Cary, NC. 21. Schuler-Malyoth, R The microorganisms of the bifidus group (Lactobacillus bifidus). III. Detection of L. bifidus in milk products. Milchwissenschaft23, Cited by L. 1.Rasic and 1.A. Kurmann, Schuler-Malyoth, R., A. Ruppert and F. Muller. 1968b. The microorganisms of the bifidus group (Lactobacillus bijidus). I. Historical review, nutritional, physiological and therapeutic aspects, morphology, cultural propertiesand taxonomy.milchwissenschaft23, Cited by L. 1. Rasic and J. A. Kurmann. (1983). 23. Smaczny, T. and M. Reinartz Joghurt, Bioghurt und Biogardeprodukte. AID Verbraucherdienst 11: Speck, M. K Preparation of lactobacilli for dietary uses. J. Food Prot. 42: Tamime, A. Y. and R. K. Robinson, Yogurt: Science and Technology. Pergamon Press, Oxford. 26. Terzaghi, B. E. and W. E. Sandine Improved medium for the lactic streptococci and their bacteriophages. Appl. Microbiol. 29: