(Eragrostis tef) fermentation. M. Ashenafi. Ersho is a clear, yellow liquid that accumulates on the surface of fermenting teff-flour batter and is collected to serve as ...
World Journal of Microbiology & Biotechnology 10, 69-73
Microbial flora and some chemical properties of ersho, a starter for teff (Eragrostis tef) fermentation M. Ashenafi Ersho is a clear, yellow liquid that accumulates on the surface of fermenting teff-flour batter and is collected to serve as an inoculum for the next fermentation. The pH of ersho samples was about 3.5 and titratable acidity ranged b e t w e e n 3.1% and 5.7%. The mean aerobic mesophilic counts from four households varied between 6.9 × 10 6 and 1.3 x 10 s c.f.u./ml and the aerobic bacterial flora consisted of Bacillus spp. Mean yeast counts ranged b e t w e e n 5.2 x 10 5 and 1.8 x 10 6 c.f.u./ml and comprised, in order of abundance, Candida milleri, Rhodotorula mucilaginosa, Kluyveromyces marxianus, Pichia naganishii and Debaromyces hansenii. Candida milleri was the m o s t dominant isolate in all samples. About 90% of the teff flour samples had aerobic mesophilic counts _~ 105 c.f.u./g and Gram-positive bacteria constituted about 71% of the total isolates. About 80% of samples had Enterobacteriaceae counts of 104 c.f.u./g.
Key words: Ersho, fermented foods, microbial flora, teff.
Enjera is a fermented, sour, leavened, pancake-like bread made from teff (Eragrostis tej), wheat, barley, sorghum or maize or a combination of some of these cereals. Teff enjera is the most common and the main staple food in much of the highlands of Ethiopia. The preparation of teff enjera consists of two stages of natural fermentation, which last for about 24 to 72 h, depending on ambient temperatures. The only required ingredients are the teff flour and water. Inoculation is accomplished by consistently using a partially-cleaned fermentation container and by adding some ersho, the clear, yellow liquid that accumulates on the surface of the batter, from a previous fermentation. This ersho contains 96.4% moisture, 0.05 mg riboflavin/100 g, and 0.4rag niacin/ 100 g (Steinkraus 1983). About 480 g ersho is added to 3 kg teff flour and 61 water. The various traditional teff threshing processes mean that teff flour probably contains a very wide variety of soil and faecal microorganisms. There have been a few studies on the microorganisms involved in teff enjera fermentation (Stewart & Asnake 1962; Gifawossen & Bisrat 1982; Gashe et al. 1982; Gashe 1985), M. Ashenafi is with the Department of Basic Sciences, Awassa College of Agriculture, Addis Ababa University, PO Box 5, Awassa, Sidamo, Ethiopia.
but only one, that of Gifawossen & Bisrat (1982) on the isolation of Candida and Pichia species, on the microflora of ersho and its importance in initiating the fermentation of teff. The present study was to identify the important microorganisms in ersho and teff flour and assess their possible role in the initiation of teff fermentation.
Materials and Methods Collection of Samples A total of 40 ersho samples were collected from four households on 10 different sampling days. The samples were microbiologically analysed within 2 h of collection.
Microbiological Analyses of Ersho Ten-ml samples of each ersho were homogenized, using a vortex mixer, before plating.
Aerobic Mesophilic Bacteria: Ersho samples were diluted in sterile water and 0.1-ml subsamples of appropriate dilutions were spread, in duplicate, on the pre-dried surfaces of Plate Count agar (PC; Merck) plates, with a bent glass rod. Colonies were counted after incubation at 30 to 32°C for 48 h.
Enterobacferiaceae: Samples (0.1 ml) of ersho were spread plated in duplicate on the pre-dried surfaces of Violet Red/Bile/Glucose/agar
© 1994 Rapid Communications of Oxford Ltd
World Journal of Microbiology & Biotechnology, Vol 10, 1994
69
M. Ashenafi
Results
(Oxoid) plates. The plates were incubated at 32°C for 24 h. Purple-red colonies were counted as Enterobacteriaceae.
Lactic Acid Bacteria: Samples (0.1 ml) of appropriate dilutions were spread-plated in duplicate on pre-dried surfaces of LSD (Oxoid) plates. Colonies were counted after incubation in an anaerobic jar (Oxoid) at 32°C for 48 h. Yeasts and Moulds: Samples (0.1 ml) of appropriate dilutions were spread-plated in duplicate on pre-dried surfaces of chloramphenicol/Bromophenol Blue/agar (CBB) consisting of (g/1 in distilled water): yeast extract, 5.0; glucose, 20.0; chloramphenicol, 0.1; Bromophenol Blue, 0.01; and agar, 15; adjusted to pH 6.0 to 6.4. Yeast colonies were counted after incubating the plates at 25 to 27°C for 5 days. Microbiological Analyses of Teff Teff samples were collected from I0 different households. Ten g of teff were homogenized in 90 ml of sterile water and appropriates dilutions were processed as the ersho, so that aerobic mesophilic bacteria, Enterobacteriaceae, lactic acid bacteria, and yeasts and moulds could be counted. Flora Assessment After colony counting, 10 to 20 colonies were selected at random from countable PC agar plates. The sub-cultures were further purified by repeated plating. A total of 400 colonies from ersho samples and 170 colonies from teff samples were differentiated into various bacterial groups. Phase-contrast microscopy was used to examine cell shape and grouping, presence or absence of endospores and motility. Gram reaction was determined using the KOH test of Gregersen (1978). Cytochrome oxidase was tested by the method of Kovacs (1956) and catalase with 3% (v/v) H202 solution. Glucose metabolism was investigated by the O/F test of Hugh & Leifson (1953). Yeast isolates from ersho samples were characterized on the basis of their fermentation of and gas production on glucose, galactose, sucrose, maltose, lactose or raffinose, according to the methods of Farkas (1984). Assimilation of these sugars and nitrate were tested by the auxanographic method of Barnett et al. (1979). Determination of pH and Titratable Acidity The pH of ersho was measured using a digital pH meter. A volume (0.9 ml) of ersho was titrated against 0.1 M NaOH to determine titratable acidity as lactic acid (Ecldes et al. 1951).
The pH values of ersho samples collected from the various households were around 3.5 and did not show any marked variations within or between households [coefficient of variation, (CV) < 6%]. The titratable acidity ranged between 3.03% and 5.7% for all samples and showed significant variation within households (CV > 25%) and within all samples (CV = 47.6%). (Table 1). The mean aerobic mesophilic counts of ersho samples varied between 6.9 x 106 and 1.3 x 108 c.f.u./ml and the mean yeast counts ranged between 5.2 x 105 and 1.8 x 106 c.f.u./ml (Table 1). Counts of Enterobacteriaceae and lactic acid bacteria were < 10 and 102 c.f.u./ml, respectively, for all samples (data not given). The aerobic mesophilic bacterial flora consisted of only Bacillus spores. Characterization of yeast isolates revealed that five species constituted most of the yeast flora. These were, in order of abundance, Candida milleri, Rhodotorula mucilaginosa, Kluyveromyces marxianus, Pichia naganishii and Debaromyces hansenii, (Figure 1). Candida milleri was isolated from 85%, 100%, 80% and 71% of the samples collected (at different times) from households A, B, C and D, respectively (Figure 2) and R. mucilaginosa was isolated from more than 40% of the samples collected from households A, B and D. Only a few samples from households B and D contained K. marxianus and only a few from households A, C and D had P. naganishii. Debaromyces hansenii was only isolated from about 30% of samples from household A. About 70% of the teff flour samples had mould counts of 10 3 c.f.u./g. About 80% had Enterobacteriaceae counts of 104 c.f.u./g and about 90% had aerobic mesophilic counts > 105 c.f.u./g (Figure 3). A total of 170 dominant bacterial strains was isolated from the teff flour samples (Table 2). Micrococcus, Aeromonas, Staphylococcus and Streptococcus species were isolated from eight, six, six and five of the 10 flour samples, respectively. Gram-positive bacteria constituted about 71% of the total isolates. Micrococcus spp., coryneforms, staphylococci and
Table 1. Some chemical properties and microbial counts (log c.f.u.lml) of ersho samples collected from four households.* Household Titratable acidity (%)
pH
A B C D All
Aerobic mesophilic bacteria
.~
SD
%CV
x
SD
%CV
,~
SD
%CV
•
SD
&CV
3.5 3.5 3.5 3.4 3.5
0.13 0.18 0.19 0.16 0.16
3.6 5.3 5.4 4.9 4.7
3.03 4.5 4.9 5.7 4.46
1.11 1.21 2.70 1.92 2.12
36.5 26.7 55.3 34.0 47.6
7.49 8.10 7.36 6.84 7.42
0.39 1.01 0.95 0.88 0.87
5.1 12.4 13.0 12.8 11.8
6.25 5.91 6.15 5.72 6.03
0,51 0.34 0.50 0.53 0.50
8,1 5.7 8.2 9.0 8.3
* Values shown are means (x-}, standard deviations (SD) and coefficients of deviation (CV).
70
Yeasts
WorldJournalof Microbiology& Biotechnology,Vol 10, 1994
Microflora of ersho 1 O0
60
50-7
80
O
J
o 40-
"5
O
o~
60
v
v
Jl (D
f
30-
J
,4
t'~
.q .q
/J J
E
/i A
$4
if)
~
",,1
,'q
e0
G \\
/t
E m
40 J~
©
..Q
4..-,
J-i
13.
/1
0
,'q ,'q
0
./1
20
10
A
B
C
D
Total
Household
!i A
A
A
.Q
Jl::~ ..d /qx~ Jl:x
/t
B
C
D
Total
Household
Figure 1. Distribution of C. milled ([~), R. mucilaginosa ([~), K. marxianus ( ~ ) , P. naganishii (r~) and D. hansenii ( I ) isolated from ersho from four households.
Figure 2. Distribution of C. milled ([q), R. mucilaginosa ([]), K. marxianus (D), P. naganishii ([~) and D. hansenii ( 1 ) in ersho samples collected from four households.
Bacillus spp. made up about 91% of the Gram-positive isolates and Aeromonas spp. constituted about 69% of the Gram-negative isolates.
variation may be one of the reasons why enjera produced at one household at different times or in different households could have different flavours. The physiological properties of the yeast isolates in this study indicated that the Candida and Kluyveromyces species were active gas producers from glucose, sucrose and a variety of other sugars. Pichia naganishii and Debaromyces hansenii produced a certain amount of gas from glucose only. In addition, all isolates except D. hansenii, which was only isolated from one household and then at very low frequency, are known not to utilize starch (Barnett et at. I979). These species could only be important in the fermentation of teff when fermentable sugars are available after the degradation of teff starch. Candida and Pichia spp. were reported to be dominant in the yellow fluid at the top of teff dough at 4 8 h (Gifawossen & Bisrat 1982). The second most frequently isolated species, R. mucilaginosa, was fermentatively inactive and may not therefore be important in leavening the batter of teff. Since ersho is the clear yellow liquid decanted from the batter at the end of the primary fermentation, the yeast isolates in this study could be important during this stage
Discussion The very low pH values of the ersho (about 3.5), consistently recorded for all samples, would be inhibitory for most kinds of microorganisms. The absence of members of Enterobacteriaceae or lactic acid bacteria, which were reported to be important in teff fermentation (Gashe et al. 1982), indicated that the ersho could not serve as a source of these bacteria. The aerobic mesophilic flora only consisted of Bacillus spores as the vegetative forms could not survive the low pH. Although five yeast species were isolated in this study, only C. milleri was ubiquitous and it may be the most important yeast species in the final stages of the primary fermentation. Interestingly, in household A, all five yeast species appeared at various frequencies, whereas in households B and C only two yeast species occurred. This
World Journal of Microbiology & Bio~:echnology, VoI I0, 1994
71
M. Ashenafi 100
of fermentation. According to Gashe (1985), the yeast count was (103 c.f.u./g until fermentation has gone on 24 h. Thus, the yeast flora isolated from ersho, even those which were fermentatively very active, may not be important as starters in the initiation of teff fermentation. Another source of inoculum for teff fermentation could be the teff flour itself. The traditional threshing processes of teff result in the contamination of the teff seeds with a very wide variety of soil and faecal material (Gashe eta]. 1982; Steinkraus 1983). Gram-positive bacteria dominated the aerobic flora of teff in this study and among these, micrococci and Bacillus spp. could be important in the initiation of the fermentation until members of Enterobacleriaceae could reach a large enough number to make any marked contribution to the fermentation, i.e. about 12 h (Gashe 1985). The high counts of Bacillus spores in ersho and Bacillus and Micrococcus spp. in teff flour may contribute markedly to the fermentation of teff flour. In the absence of acid-forming lactic acid bacteria, micrococci may acidify the flour-and-water paste or Bacillus may grow, producing lactic acid, gas, alcohol, acetoin and small amounts of esters and aromatic compounds (Anon. 1980). Controlled experiments are, however, needed in order to determine the actual role of Micrococcus and Bacillus spp. in the initiation of teff fermentation. The low numbers of moulds in the teff examined in this study (103 to 104/ml) are in agreement with the reports of Hobbs & Greene (1976). Moulds, however, may not be important in such acidic fermentations.
-
80
L'K'K
2
~:::,
60
iiii !i i!
"5
,,-,,, ,N'q
~-
,NN \\\
!Z! IZ
co
© ~.
4O
E
J/
\\~ \\\
03
x-x\
iiii
20
",'3, "'3, 103
10 4
Count
10 e
10 5
(c.f.u./ml)
Figure 3. Frequency distribution of aerobic mesophilic bacteria (r~), m e m b e r s of Enterobacteriaceae ([~) and moulds ( D ) in teff flour samples.
Acknowledgements The technical assistance of H. Alemayehu and T. Bekele is acknowledged.
References Table 2. Dominant bacterial isolates from the teff flour samples. Bacterial spp.
Positive samples
Isolates No.
%
Gram-positive
Bacillus Coryneforms Micrococcus
Staphylococcus Streptococcus
3 4
24 28
14.1 16.5
8
31
18.2
6 5
27 11
15.9 6.4
2 6 3 1 1
4 34 7 2 2
2.4 20.0 4.1 1.2 1.2
10
170
Gram-negative
Acinetobacter Aeromonas Alcaligenes Enterobacteriaceae Pseudomonas Totals
72
World]ournaIof Microbiology& Biotechnology,Vol I0, 1994
100
Anon. 1980. Microbial Ecology of Foods, Vol. II, Food Commodities. New York: Academic Press. Bamett, J.A., Payne, R.W. & Yarrow, D. 1979 A Guide to Identifying and Classifying Yeasts. Cambridge: Cambridge University Press. Eckles, C.H., Combs, W.B. & Macy, H. 1951 Milk and Milk Products. Bombay, New Delhi: Tata McGraw Hill. Farkas, J. 1984 Microbiological Procedures. In Testing Methods in Food Microbiology, ed. Kiss, I. pp. 28-108. Amsterdam: Elsevier. Gashe, B.A. 1985 Involvement of lactic acid bacteria in the fermentation of tef (Eragrostis teJ), an Ethiopian fermented food. Journal of Food Science 50, 800-801. Gashe, B.A., Girma, M. & Bisrat, A. 1982 Tef fermentation. 1. The role of microorganisms in fermentation and their effect on the nitrogen content of tef. SINET, Ethiopian Journal of Science 5, 69-76. Gifawossen, C. & Bisrat, A. 1982 Yeast flora of fermenting tef (Eragrostis teJ) dough. SINET, Ethiopian Journal of Science 5, 21-25. Gregersen, G. 1978 Rapid method for distinction of Gram negative from Gram positive bacteria. European Journal of Applied Microbiology 5, 123-127.
Microflora of ersho Hobbs, W.E. & Greene, V.W. 1976 Cereal and Cereal Products. In
Compendium of Methods for the Microbiological Examination of Foods, ed Speck, M.L. pp. 509-607. Washington DC: American Public Health Association. Hugh, R. & Leifson, E. 1953 The taxonomic significance of fermentative versus oxidative Gram negative bacteria. Journal of Bacteriology 66, 24-26. Kovacs, N. 1956 Identification of Pseudomonas pyocyanae by the oxidase reaction. Nature 178, 703.
Steinkraus, K.H. 1983 Handbook of Indigenous Fermented Foods. Microbiological Series, Vol. 9. New York: Marcel Dekker, Inc. StewarL R. & Asnake, G. 1962 Investigations of the nature of injera. Economic Botany 16, I27-130.
(Received in revised form 15 June 1993; accepted 21 June 1993)
World Journal of Microbiology & Biotechno]ogy, Vol 10, I994
73
