LWT - Food Science and Technology 155 (2022) 112927

Contents lists available at ScienceDirect

LWT
journal homepage: www.elsevier.com/locate/lwt

Fermentation dynamics of spontaneously fermented Ethiopian honey

wine, Tej
Eskindir Getachew Fentie a, b, c, Minsoo Jeong b, Shimelis Admassu Emire c, Hundessa
Dessalegn Demsash c, Min A. Kim b, d, Jae-Ho Shin b, *
a
College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, Addis Ababa, 16417, Ethiopia
b
Department of Applied Biosciences, Kyungpook National University, Daegu, 41566, South Korea
c
School of Chemical and Bio-Engineering, Addis Ababa Institute of Technology, Addis Ababa University, P.O. Box 385, King George VI Street, Addis Ababa, 16417,
Ethiopia
d
National Institute for Korean Medicine Development, Daegu, 38540, South Korea

A R T I C L E I N F O A B S T R A C T

Keywords: Spontaneously fermented Ethiopian honey wine, Tej, is a popular alcoholic beverage in Ethiopia. However, a
Tej detailed information which is needed to modernized the production process of this traditional beverage has not
Dynamics yet been conducted. Thus, this paper is aimed to reveal the physicochemical and microbial dynamics occurred
Microbiota
during Tej fermentation processes. To address this objective, Tej samples was collected every 48 h from three
Physicochemical
Antioxidant
households located at three different cities of Ethiopia. As fermentation progressed, pH, glucose, and fructose
content decreased, whereas, titratable acidity and alcohol content increased. Likewise, antioxidant activity,
especially 3-ethylbenzothiazoline-6-sulfonic acid diammonium salt (ABTS) assay, increased as the fermentation
progressed to final stages. Moreover, the bacterial communities shifted to Lactobacillus dominancy as the
fermentation progress to the end. Furthermore, during the whole fermentation period, fungal communities were
practically dominated by Saccharomyces species. Due to these dominances microbial diversity decreased toward
the end of fermentation. Besides, quantitative increases in both microbial communities were observed during Tej
fermentation periods. However, fungal gene copy numbers were by half higher than bacterial gene copy
numbers. In conclusion, Species of Lactobacillus and Saccharomyces could be taken as a potential mixed starter
culture candidates for inoculated Tej fermentation systems.

spontaneous fermentation systems (Bahiru et al., 2006).

Crude honey and “gesho” (Rhamnus prinoides) together with natu­
1. Introduction
rally occurring microorganisms are the basic ingredients of good quality
Tej (Lemi, 2020). The process begins by mixing honey and water in a 1:3
In Ethiopia, traditional fermented foods and beverages are produced
ratio. Then, after 2–3 days of primary fermentation, the blend is filtered
in households and consumed in significant quantities (Fentie et al.,
using a clean cheesecloth. Then, boiled “gesho” leaves and stems are
2020). Traditional Ethiopian alcoholic beverages have at least an
added to the honey water filtrate. At this point, some communities use
equivalent domestic market share as commercial alcoholic beverages
fresh “gesho” leaves instead of dried and boiled leaves and stems.
(Mulaw & Tesfaye, 2017). The spontaneous fermentation process makes
Further, some producers add small amounts of malt powder to the boiled
these beverages to have acid alcoholic properties (Nemo & Bacha,
“gesho” leaves (Nemo & Bacha, 2021). Regardless of the ingredients, the
2020). Tej is one such commonly consumed fermented alcoholic
blend undergoes a secondary fermentation for 8–21 days. Finally, the
beverage primarily made from crude honey (Bahiru et al., 2001).
mixture is filtered and served as the final product, Tej.
Approximately 80% of honey produced in the country is used to produce
Good quality Tej is cloudy in appearance, yellow in color, and sweet
this beverage (Gebremedhin et al., 2013). As a matter of fact, Saccha­
in flavor (Vogel & Gobezie, 1995). Unfortunately, it is often difficult to
romyces cerevisiae inoculated fermentation systems are currently used to
achieve these key quality characteristics due to the spontaneous nature
produce honey wine in other parts of the world (Balogu & Towobola,
of the fermentation (Lemi, 2020). To generate consistent product
2017). However, traditional Ethiopian honey wine is still produced via

* Corresponding author.
E-mail address: jhshin@knu.ac.kr (J.-H. Shin).

https://doi.org/10.1016/j.lwt.2021.112927
Received 26 August 2021; Received in revised form 7 November 2021; Accepted 1 December 2021
Available online 3 December 2021
0023-6438/© 2021 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
E.G. Fentie et al. LWT 155 (2022) 112927

Ababa (AA), Bahir Dar (BD), and Debre Markos (DM), Ethiopia (Fig. 1).
Abbreviations From the first to the last day of fermentation, samples were aseptically
collected every 48 h and transported to the laboratory in insulated ice
TA – Titratable acidity boxes plus freezer packs. Samples not requiring immediate analysis were
AA household Tej sample from Addis Ababa stored at − 20 ◦ C.
BD household Tej sample from Bahir Dar
DM household Tej sample from Debre Markos 2.2. Physicochemical analysis

2.2.1. pH and titratable acidity (TA)

Sample pH and TA levels were measured using standard methods
quality, the natural fermentation process must be substituted with a
(OIV, 2020). All measurements were conducted in triplicate.
direct inoculation fermentation system (Steinkraus, 2004), therefore
understanding the dynamic behaviors of the system are key toward
2.2.2. Sugar profiling
achieving this goal. Thus, unraveling microbial communities and exo­
Sample glucose, fructose, and sucrose levels were quantified using
metabolomic profile dynamics will definitely help to get a full picture of
high a performance liquid chromatography-evaporative light scattering
the spontaneous fermentation system (Tamang & Thapa, 2006).
detector (HPLC-ELSD) system (Shimadzu, Japan) (Ma et al., 2014). A 5
Previous researches looked into the physicochemical and microbio­
mL Tej sample was centrifuged at 3200 rpm for 30 min and the super­
logical aspects of Tej samples obtained from traditional alcohol retailers
natant double filtered through 0.45 μm and 0.22 μm pore size mem­
in the Addis Ababa and Jima provinces of Ethiopia (Bahiru et al., 2001,
branes. A LC-20AT HPLC, YMC-Pack Polyamine II (250 × 4.6 mm. D.S-5
2006; Nemo & Bacha, 2020). However, most of these studies focused on
final product characterization. Nevertheless, the final product is the
μm, 12 nm) column and ELSD-LT II detector were used for analyses. A
10 μL aliquot was injected onto the HPLC-ELSD at a column temperature
result of combined multiple dynamic interactions between microbes and
of 30 ◦ C. An isocratic mobile phase of acetonitrile: water at a 75:25 ratio
produced metabolic products during the fermentation process (Weckx
with flow rate of 1.0 mL/min was used for analysis. Nitrogen (N2) was
et al., 2010). Hence, final product characterization alone does not un­
used as a nebulizer gas at a 2 L/min flow rate. Finally, sugar quantifi­
cover the complete story of the spontaneous Tej fermentation system. To
cation was conducted by comparing samples sugar peak areas with the
the best of our knowledge, there is a dearth of information in the liter­
corresponding standards calibration curves.
ature particularly focused on Tej fermentation dynamics. To address this
knowledge gap, we investigated changes in physicochemical properties,
2.2.3. Ethanol quantification
antioxidant activities, and microbial communities over time during the
Ethanol level was determined by following the standard method
spontaneous fermentation of Ethiopian honey wine.
developed by OIV (2020). A 200 mL of Tej aliquot was added to a
volumetric flask. Then, the measuring cylinder was rinsed by distilled
2. Material and methods
water of three 20-mL aliquots. A 20 mL of distilled water was added to
another 200 mL volumetric flask that used to collect the distillate. This
2.1. Sample collection and transportation
collection flask was placed in a cold-water bath throughout the whole
distillation period. Distilled water was then filled to the mark of distil­
Tej samples were collected from three different households in Addis
late collection volumetric flask. Finally, the alcohol content was

Fig. 1. Area map for a time series of Tej samples from three Ethiopian households. The left side represents Ethiopia, while the right-side represents an enlarged map
of sampling locations. The red dot, star and diamond shapes represents AA, DM and BD household samples. (For interpretation of the references to color in this figure
legend, the reader is referred to the Web version of this article.)

2
E.G. Fentie et al. LWT 155 (2022) 112927

calculated using a standard table of aqueous organic solutions. 95 ◦ C for 30 s, 58 ◦ C for 30 s, 72 ◦ C for 30 s, and 72 ◦ C for 5 min (Jung
et al., 2020). After checking library quality and quantity, it was sub­
2.3. Antioxidant capacity jected to Illumina Miseq sequencing as described in manufacturer’s
instructions.
2.3.1. DPPH (2,2-diphenyl2,2-diphenyl-1-picrylhydrazyl (DPPH) assay
The DPPH assay was conducted according to Blois (1958). A 2 mL 2.4.4. Quantitative PCR
samples were first centrifuged at 14900 rpm to remove suspended par­ Quantitative PCR (qPCR) was performed according to Park et al.
ticles. Then, 20 μL sample supernatant was mixed with 150 μM DPPH (2009) with minor modifications. First, Escherichia coli DH5α was
(D9132-Sigma Aldrich, USA) and 180 μL ethanol solution. Concurrently, transformed using the one copy ITS and 16SrDNA fragment pTOP TA V2
a blank solution was prepared by dissolving 180 μL DPPH in 20 μL cloned vector. Plasmid DNA was then extracted using the GeneJET
ethanol. After a 15 min incubation period at 37 ◦ C, the optical density Plasmid Miniprep kit (Thermo Scientific, USA) following manufacturer’s
(OD) was measured at 517 nm (TECAN, Spark 10M Multimode Micro­ instruction. This DNA was then serially diluted; 101–109, with con­
plate Reader, Switzerland). Finally, DPPH activity was calculated as centrations measured using a Qubit 2.0 fluorometer (Life Technologies).
follows; A 10 μL PCR reaction mixture comprised 5 μL MG 2X qPCR Master mix
(SYBR Green), 1 μL DNA template, 0.3 μL universal forward primer
DPPH activity (%) = (Blank OD − Sample OD / Blank OD) × 100
(Table 1), 0.3 μL universal reverse primer (Table 1) and 3.4 μL SDW. The
following qRT–PCR conditions were used; 95 ◦ C for 5 min followed by
2.3.2. ABTS (3-ethylbenzothiazoline-6-sulfonic acid diammonium salt 40 cycles of 95 ◦ C for 30 s, 57 ◦ C (bacteria)/62 ◦ C(fungi) for 30 s, and
(ABTS) assay 72 ◦ C for 30 s. Standard curves for serially diluted plasmid DNA were
The ABTS assay was performed according to Re et al. (1999)). First, generated. and gene copy numbers for bacteria and fungi were
an ABTS cation radical solution was formed by mixing 7 mM ABTS with calculated.
2.45 mM potassium persulfate. After a 12–16 h incubation in darkness,
the solution was diluted in 5 mM phosphate buffer to an absorbance of 2.4.5. Bioinformatics analysis
0.7 ± 0.02 at 734 nm. Then, 197 μL of this solution was mixed with a 3 Quantitative Insights into Microbial Ecology 2 (QIIME2) version (v).
μL aliquot of centrifuged Tej sample. Finally, OD values were measured 2020.11 was used to analyze raw single end reads of a trim length 250
and ABTS activity expressed as follows; base pairs (Bolyen et al., 2019). The minimum number of reads was then
ABTS activity(%) = (Blank OD − Sample OD / Blank OD) × 100 set to 1 for the purpose of quality filtering. Then, amplicon sequence
variants (ASV) were obtained using DADA2 (Callahan et al., 2016).
2.4. Microbiome analysis Subsequently, taxonomic identification was performed using the
classify-sklearn classifier for 99% similarity against UNITE and SILVA
2.4.1. DNA extraction 132 reference databases for fungi and bacteria, respectively. Similarly,
Microbial DNA from Tej samples was extracted using the QIAamp sepp and mafft algorithms were used to generate phylogenetic trees for
PowerSoil Pro kit (QIAGEN, Germany) following manufacturer’s in­ bacteria and fungi, respectively. Sequences were finally rarefied to
structions. DNA concentrations were measured using a Qubit 2.0 Fluo­ normalize differences in sequencing depths.
rometer (Life Technologies, Carlsbad, CA, USA). Good quality DNA
templates were stored at − 20 ◦ C for further analysis. 2.5. Statistical analysis

2.4.2. 16S rRNA sequencing After checking for normality and homoscedasticity (Levene’s test) of
Amplification of hypervariable region V4–V5 of 16SrRNA gene was physicochemical data, statistical significance (P < 0.05) of the tripli­
performed by using specific forward and reverse primers (Table 1). cated data was determined using one way analysis of variance in
Polymerase chain reaction (PCR) (Mastercycler Nexus GSX1, Eppendorf, RStudio 4.0.3. Microbiome data visualization and statistical analyses
Germany) amplification conditions were; 95 ◦ C for 5 min, then 15 cycles were performed in Web-based Calypso and RStudio. We determined
of 95 ◦ C for 30 s, 60 ◦ C for 30 s, 72 ◦ C for 30 s, and 72 ◦ C for 5 min (Jung α-diversity indices using the Vegan package (Oksanen et al., 2017) and
et al., 2020). Then, barcoded, amplified, and pooled DNA’s were puri­ β-diversity analyses were performed using Unifrac distance
fied using the QIAquick gel extraction kit (QIAGEN). Amplicon library measurements.
quality and quantity were then checked using the Agilent 2100 Bio­
analyzer (Agilent Technologies Inc., USA). The amplicon library was 3. Result and discussion
finally subjected to the Illumina MiSeq (Illumina-MiSeq-USA) platform
following manufacturer’s instructions. 3.1. Physicochemical property dynamics

2.4.3. Internal transcribed spacer (ITS) sequencing Acidity is one of the most important physicochemical properties that
Fungal community structures were analyzed using ITS amplicon requires monitoring during beverage production, particularly alcoholic
sequencing (Turenne et al., 1999). PCR amplification using ITS primers beverages (Holzapfel, 2015). This parameter influences the overall
(Table 1), was conducted as follows; 95 ◦ C for 5 min, then 30 cycles of fermentation process, sensorial attributes, and shelf-stability of the final

Table 1
List of primers used for amplicon sequencing and quantitative real-time PCR (qRT–PCR) amplification.
Primers purpose Target Designation Sequence (5–3) References

Amplicon sequencing Bacteria 515F GTGNCAGCMGCCGCGGTAA Parada et al. (2016)

907R CCGYCAATTYMTTTRAGTTT
Fungi ITS1 CTTGGTCATTTAGAGGAAGTA Gardes and Bruns (1993)
ITS2 GTGAATCATCGAATCTTTGAA Turenne et al. (1999)
qRT–PCR amplification Bacteria Eub338 ACTCCTACGGGAGGCAGCAG Lane (1991)
Eub518 ATTACCGCGGCTGCTGG Muyzer et al. (1993)
Fungi Fu18S1 GGAAACTCACCAGGTCCAGA Borneman and Hartin (2000)
Nussu1536 ATTGCAATGCYCTATCCCCA

3
E.G. Fentie et al. LWT 155 (2022) 112927

product (Tôrres et al., 2011). Regardless of sample collection household, < 0.05) variation on the final pH and TA values were observed for the
the acidity of all Tej samples significantly increased (P < 0.05) as samples collected from different households. Particularly, this was
fermentation progressed to the end; mean pH values fell from 3.58 to especially true for the samples collected from DM area households
2.98, 3.78 to 3.1, and 4.16 to 3.7 for AA, BD and DM household samples, (Fig. 2A and B). This may have been due to variations in raw materials,
respectively (Fig. 2A). Likewise, as fermentation progressed, the mean processing conditions, and microbial quantities and compositions dur­
TA for samples increased from 0.23 to 3.00, 0.23 to 4.01, and 0.38 to ing spontaneous fermentations.
2.25 for AA, BD and DM samples, respectively (Fig. 2B). The main source Throughout the entire fermentation period, the medium was domi­
of increased acidity levels during spontaneous alcoholic fermentation nated by monosaccharide sugars of glucose and fructose (Fig. 2C and D).
was the continuous generation of organic acids, particularly lactic acid Sucrose levels for all samples were below detectable limits even at the
formation from Lactobacillus contributes a lot for this result (Chidi et al., start of the fermentation. Mature honey, which is usually used as the raw
2018). A similar acidity increase was also observed during laboratory material for Tej, could be the primary cause for the low sucrose levels
scale Tej fermentation (Nemo & Bacha, 2021). Due to the production of (Belay et al., 2017). The mean glucose levels was significantly decreased
lactic acid by Lactobacillus species, a steeper gradient of acidity increase (P < 0.05) from 7.32 to 2.21 g/L, 8.36 to 2.80 g/L, and 7.98 to 1.70 g/L
was observed, particularly at the onset of fermentation (Fig. 2A and B). during the fermentation period for AA, BD and DM samples, respectively
Furthermore, honey’s lower buffering capacity may exacerbate this (Fig. 2C). Similarly, during fermentation, the mean fructose levels was
larger acidity gradient at the beginning of fermentation even for a lower significantly decreased (P < 0.05) from 10.24 to 4.2 g/L, 10.26 to 5.1
organic acid concentration (Tôrres et al., 2011). Although similar trend g/L, and 9.60 to 3.8 g/L for AA, BD and DM household samples,
of increase in acidity was observed for all of Tej samples, a significant (P respectively (Fig. 2D). This decreased sugar level is basically due to

Fig. 2. Progressive changes in A) pH, B)

titratable acidity, C) glucose levels, D) fruc­
tose levels, and E) ethanol levels during the
spontaneous fermentation of Tej. Each dot in
the line graph represents sugar and ethanol
levels for a Tej sample from Addis Ababa
(Green), Bahir Dar (Orange), and Debre
Markos (Violet) areas. The stars (*)in the
figure is to indicate the time when “gesho”
was added into the fermentation medium. .
(For interpretation of the references to color
in this figure legend, the reader is referred to
the Web version of this article.)

4
E.G. Fentie et al. LWT 155 (2022) 112927

biomass growth and exo-metabolite production (Pereira et al., 2017). significantly increased (P < 0.05) from 27.85% ± 2.19%–49.97% ±
However, at the start of the fermentation, glucose uptake rates were 1.06% for DPPH and 15.77% ± 2.18%–49.69% ± 1.06% for ABTS assay
much faster than fructose consumption (Fig. 2C and D); specifically, 3- (Fig. 3A and B). Ascorbic acid, α-tocopherol, catalase, β-carotene, and
and 2-fold decreases in glucose and fructose levels were observed, peroxides are all antioxidant activity sources in honey (Chua et al.,
respectively. The microorganisms involved in Tej fermentation might 2013). Besides, Chen et al. (2020) reported good antioxidant capacity
prefer glucose over fructose, which could be the possible reason for this for “gesho”. Therefore, these two raw materials are major antioxidant
result. Similar glucose and fructose uptake rates were also observed for activity sources for Tej samples. The progressive increase in Tej antiox­
honey wine fermentation using different nitrogen supplement concen­ idant capacity was rapid between days 3 and 12 of fermentation period
trations (Cristian et al., 2015). In our study, after the day 16, glucose (Fig. 3). This was most probably due to the extraction of bioactive
consumption rates were decreased and then plateaued (Fig. 2C and D). compounds from “gesho”. After 12 days, both antioxidant activity assays
This may have been due to adverse conditions caused by primary and displayed almost constant values until the last day of fermentation
secondary metabolites produced by fermentative microbes (Iglesias (Fig. 3). Comparable antioxidant capacity results were previously
et al., 2014). Especially the increase in ethanol content, which is a observed for cornelian cherry fruit juice supplemented honey wine
known microbial growth inhibitor, might have a lion-share contribution (Adamenko et al., 2018).
for the result. Furthermore, in terms of household comparisons, BD A gentle and high increases were observed for DPPH and ABTS
samples exhibited comparatively higher rates of sugar consumption, values, respectively (Fig. 3). This variation may have been attributed to
whereas, DM samples exhibited a gentler sugar consumption throughout the lower water solubility of DPPH radicals (Arnao, 2001). Meanwhile,
the fermentation period (Fig. 2C and D). Household differences were for Tej samples from different households, time-dependent antioxidant
primarily caused by the difference in microbial composition inherited activity variations were observed (Fig. 3). For instance, after the third
from spontaneous fermentation of Tej (Marshall & Mejía-Loríio, 2011). day of fermentation, DM samples showed a significant increase (P <
The major objective of an alcoholic beverage fermentation process is 0.05) in ABTS levels. Thus, how “gesho” was added to the fermentation
to convert available sugars to ethanol via fermentative microorganisms medium, environmental factors, different floral sources, and prepara­
(Buglass, 2010). Thus, intrinsic biochemical reactions require conducive tion and processing conditions may have contributed to
environments to achieve this phenomenon (Steinkraus, 2004). The sit­ household-based antioxidant capacity variations (Akalın et al., 2017).
uation may become more exacerbated for a carbon source such as honey,
which has a lower pH, low buffering capacity, high osmotic pressure,
3.3. Quantitative microbial dynamics
and deficiencies in essential nutrients (Pereira et al., 2017). However, in
this study, all our Tej samples showed progressive alcohol level in­
The qualitative analysis of microbial comunity structures must be
crements especially after the 8th fermentation day (Fig. 2E). This late
start in alcohol production may have been due to an extended lag supported by quantitative analyses to generate a complete view of mi­
crobial kinetics during spontanous fermentation (Kim et al., 2015). In
growth phase in microbial growth rate (Aranda et al., 2019). A
maximum alcohol content (12.02 ± 0.02 g/100 mL) was recorded after this study, we used qPCR for the absolute quantification of total bacteria
and yeast levels during Tej fermentations. Serially diluted pure genome
24 days of fermentation in BD samples; besides, alcohol production rate
during the fermentation period were also higher for this sample were used to develope a standard curve having an equation of Y =
(Fig. 2E). The dominancy by Lactobacillus, Zymomonas, and Saccharo­ − 3.5261X + 4.7208, R = 0.9954 for fungi, and Y = − 2.267X + 12.951,
myces in this particular sample may have contributed to this observation R = 0.9877 for bacteria (Y = cycle threshold; X = log10 DNA concen­
(Fig. 4). However, after the 20th day of fermentation, ethanol produc­ tration at ng/μL).
tion in all samples declined, then plateaued (Fig. 2E). The inactivation Quantitatively, total bacteria and fungi increased during fermenta­
and/or death of microbes at this stage of fermentation might cause this tions (Fig. 4A). The gene copy numbers of total bacteria increased from
kind of result (Martínez-Moreno et al., 2012). 2.92 × 107–1.25 × 1010, 2.53 × 105–1.53 × 1010, and 2.56 × 105–7.5 ×
109 ng/μL for AA, BD, and DM household samples, respectively. Like­
wise, the gene copy numbers of total fungi increased from 9.5 ×
3.2. Antioxidant activity dynamics 106–1.91 × 1010, 4.29 × 106–2.68 × 1010, and 4.16 × 106–1.1 × 1010
ng/μL for AA, BD, and DM household samples, respectively (Fig. 4A).
In this study, both colorimetric based antioxidant assays (DPPH and The growth of Lactobacillus and Saccharomyces fermentative microbes
ABTS (+) were used to monitor the development of functional property underpinned these sharp increments in total bacterial and fungal gene
of Tej. During fermentation periods, cumulative antioxidant activities copy numbers (Fig. 4B and C). A previous study using microbial

Fig. 3. Dynamics of A) DPPH and B) ABTS levels during Tej fermentation. Each differently colored line represents antioxidant activity values for Tej sample collected
from Addis Ababa (Green), Bahir Dar (Orange), and Debre Markos (Violet) areas. The stars (*)in the figure is to indicate the time when “gesho” was added into the
fermentation medium. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

5
E.G. Fentie et al. LWT 155 (2022) 112927

Fig. 4. Time series changes of A) bacterial and fungal gne copy number, relative aboundance of B) bacteria and C) fungal communities during Tej fermentaiton.
Samples were from Addis Ababa, Bahir Dar, and Debre Markos, Ethiopia. Each dot in the line graph represent ITS (Orange) and 16SrDNA (Violet) gene copy numbers
during the fermentation period. The stars (*)in the figure is to indicate the time when “gesho” was added into the fermentation medium. (For interpretation of the
references to color in this figure legend, the reader is referred to the Web version of this article.)

cultivation methods reported similar increments of lactic acid bacteria processes. Nevertheless, both qualitative and quantitative data revealed
and yeast from 4.01 to 8.88 and 4.16–9.41 log CFU/mL during Tej fer­ that Lactobacillus and Saccharomyces genera were compatible and per­
mentations, respectively (Nemo & Bacha, 2021). forming well during successful Tej fermentation processes (Fig. 4). Be­
After the day 10 of fermentation, the gene copy numbers of both sides, bacteria to fungi ratios indicated that bacterial levels were almost
bacteria and fungi exhibited almost constant values across all household 50% lower than fungal levels. The stressful environmental conditions in
samples (Fig. 4A). This situation more than likely reflected microbial honey may have induced the rapid growth of Saccharomyces over
flora stationary growth phases; limited essential nutrients, increased cell Lactobacillus. Similar data were also observed during the alcoholic
to cell interactions, and toxic metabolite accumulation may have fermentation of Brazilian Cauim (Ramos et al., 2010).
induced microbes to entry into this phase (Jaishankar & Srivastava,
2017). A similar pattern was observed for bacteria and fungi gene copy 3.4. Microbial community dynamics
numbers during the fermentation of Korean kimchi (Park et al., 2009). In
our study, bacterial to fungal ratios during stationary phases were 0.6, Microbial community structures in Tej fermentations were domi­
0.5, and 0.7 for AA, BD, and DM household samples, respectively nated by particular groups of bacterial and fungal species (Fig. 4). The
(Fig. 4A). continuous production of microbial metabolites, high sugar levels,
A small quantitative variation in different household samples was competition between microbes and anaerobic conditions during
most probably caused by the spontaneous nature of fermentation fermentation processes may have contributed to this dominance

6
E.G. Fentie et al. LWT 155 (2022) 112927

(Holzapfel, 2015). Thus, as fermentation progressed, only microorgan­ microbe for the conversion of sugar to ethanol. Thus, the predominancy
isms withstanding competition between microorganisms and unfavor­ of Saccharomyces during the fermentation of alcoholic beverages is the
able conditions dominated and survived until fermentation end. usual phenomenon encountered during successful alcoholic fermenta­
Bacterial communities during Tej fermentation were dominated by tion (Walker & Stewart, 2016). Other Monascus, Zygosaccharomyces,
the Firmicutes and Proteobacteria phyla (Fig. 4B). Firmicutes were by far Candida, and Torulaspora species were also detected at very low relative
the most abundant phylum (>80%), especially near fermentation end. abundance levels (Fig. 4C) at early fermentation stages, but were later
Proteobacteria was the next most abundant phylum, particularly at early dominated by Saccharomyces (Fig. 4C). A similar Saccharomyces domi­
and/or fermentation end periods (Fig. 4B). Lactobacillus, Clos­ nant status was observed during the spontaneous fermentation of Cat­
tridium_sensu_stricto_1, Romboutsia, Paeniclostridium, Peptos­ alanesca grapes (Di Maro et al., 2007).
treptococcaceae, Lachnospiraceae, Paenibacillus, Bacillus, Thus, Lactobacillus and Saccharomyces were the dominant genera
Enterococcaceae, and Oceanobacillus were the observed families and during Tej fermentation. The traditional wineries microbiota in each
genera under the phyla of Firmicutes. Similarly, Zymomonas, Acetobacter, household is believed to have a big contribute for these exclusive
Pseudomonas, Comamonadaceae, and Enterobacteriaceae were the dominancy of certain microbes. Indeed, both are the most common
observed genera and families under the phyla of Proteobacteria (Fig. 4B). microbes in several spontaneously fermented alcoholic beverages (Jung
As a matter of fact, with expected difference in relative abundance these et al., 2012). Furthermore, most species in these genera are under the
two phyla are the most expected bacterial phyla in fermented beverages category of “generally recognized as safe (GRAS)” (Vida-Plavec & Ber­
(Cason et al., 2020). A similar dominance was reported in other tradi­ lec, 2020).
tional fermented alcoholic beverages (Diaz et al., 2019). Throughout the
fermentation period, fungal communities were exclusively dominated 3.5. b3.5 microbial diversity dynamics
by the Ascomycota phylum. Saccharomyces, Candida, Torulaspora, and
Zygosaccharomyces were the observed genera under the phyla of Asco­ Bacterial and fungal α-diversity analyses were performed to char­
mycota (Fig. 4C). This phylum is the most numerous and diverse fungal acterize microbiome diversity changes during Tej fermentation. The
group (Taylor et al., 2015). A similar Ascomycota dominance was α-diversity indices, namely, Chao1, Shannon, Evenness, Simpson, and
observed in Indian alcoholic beverage starters (Sha et al., 2019). InvSimpson, were calculated for all household samples (Table 2, Fig. 5A
Lactobacillus was the major bacterial genus observed during Tej and B). The Chao1 index was used to demonstrate species richness dy­
fermentation (Fig. 4B). Its relative abundance increased from 24% to namics occurred during the fermentation of Tej. Irrespective of collec­
100%. The species of Lactobacillus has a good capability to produce tion households, both bacterial and fungal community richness had
compounds that contribute for the development of distinct favors in decreased as the fermentation neared completion (Table 2). The domi­
alcoholic beverages (Fuochi et al., 2019). Although the exclusive nancy of particular microbes could be the primary reason for the pro­
abundance of Lactobacillus was observed in all of the samples, increment gressive decrease in Chao1 index. Furthermore, microbial diversity
patterns showed variations among household samples (Fig. 4B). Simi­ changes in bacteria and fungi were analyzed using Shannon and Simp­
larly, Lactobacillus reportedly dominated the final fermented Tej product son indices (Table 2, Fig. 5A and B). Progressive and sharp decreases
and the Korean traditional beverage, Shindari (Bahiru et al., 2006; Cha were observed for bacterial and fungal communities, respectively.
et al., 2014). However, a particular increase in bacterial community diversity was
The Zymomonas genus was also observed during Tej fermentations observed the day 4 of fermentation. The addition of “gesho” and
(Fig. 4B). A higher relative abundance of this genus was detected at day anaerobic fermentation conditions may have caused this observation.
4 of fermentation which may have been due to “gesho” addition on day Evenness was also used to demonstrate each species numbers in a given
3. However, after 10th day of fermentation, BD samples displayed a sample (Haig et al., 2015). Both bacterial and fungal communities
19.4% Zymomonas relative abundance, whereas AA and DM samples showed a lower evenness during Tej fermentation process. As fermen­
were exclusively dominated by Lactobacillus (Fig. 4A). In fact, species of tation progressed, Lactobacillus and Saccharomyces dominancy may have
this genus, in particular Zymomonas mobilis, are frequently observed been a reason for this low evenness value.
species in traditional alcoholic beverages (Diaz et al., 2019). This could A Principal Co-ordinate Analysis (PCoA) unweighted UniFrac plot
be due to the ability of the species to produce higher alcohol levels from was used to demonstrate a β–diversity of microbial communities during
mono- and disaccharide sugars (He et al., 2014). Tej fermentation (Fig. 5C and D). The plot basically demonstrates the
The Acetobacter genus was observed at the beginning, day 4, and/or level of difference in microbial tax during the fermentation period for
at fermentation end (Fig. 4B). This bacterium degrades ethanol to acetic the samples collected from those three households. The PCoA plot for
acid in the presence of oxygen (Bartowsky et al., 2003). The relative bacterial communities showed a visible cluster and a move to one di­
abundance of this genus was decreased from 1.0% to 0.0%, 10.8%– rection as the fermentation period gradually ended. A visible cluster
0.1%, and 4.0%–1.0% during fermentation periods in AA, BD, and DM distance (P < 0.0025), with a cumulative variance of 77% was observed
samples, respectively (Fig. 4B). These decreases in relative abundance for bacterial communities as the samples were compared based on the
were probably caused by a limited access to ethanol and oxygen at the fermentation time (Fig. 5C). This is because, as the fermentation period
beginning and during fermentation periods, respectively (Motlhanka advanced to the end, bacterial communities were dominated by Lacto­
et al., 2018). Additionally, the competition from fermentative micro­ bacillus. Additionally, AA and DM household samples had shown dif­
organisms might also contribute for this result. These decreases also ference visible cluster of distance form BD household samples. Similarly,
indicated that the source of contamination might be during preparation fungal communities showed a visible cluster distance (P < 0.0146) in the
and/or addition of raw materials. PCoA plot, with a cumulative variance of 95.6% for a time series Tej
Pathogenic Pseudomonas and Enterobacteriaceae microbes were also samples (Fig. 5D). Like bacterial communities, fungal communities were
observed at the raw material addition step (Fig. 4B). However, they were moving in one direction as the fermentation period was ending. The
not observed in any samples after the day 6 of fermentation. This was dominant Saccharomyces majorly contributed to this specific observa­
due to unfavorable conditions created by metabolic products from tion. Similarly, there was no visible cluster of distance observed for
dominant fermentative microorganisms. In a previous study using qualitative analysis of fungal taxes based on household comparisons.
traditional microbial plating techniques, a similar decrease in pathogens Generally, the time dependent β-diversity analyses demonstrated both
was also observed during the Tej fermentation period (Nemo & Bacha, bacterial and fungal communities moved in one direction as the
2021). fermentation gradually ended (Fig. 5C and D).
Saccharomyces was the dominant fungal genus during the whole Tej
fermentation period (Fig. 4C). In fact, this genus is the main responsible

7
E.G. Fentie et al. LWT 155 (2022) 112927

Table 2
Alpha diversity indices of time dependent Tej samples collected from three households.
Bacterial communities alpha diversity

Indices Locations Fermentation days

0 2 4 6 8 10 12 14 16 18 20 22

Chao1 AA 58 66 53 30 1 1 1 1 2 1 1 1
BD 40 14 44 13 16 14 13 13 14 19 14 15
DM 57 24 41 20 10 1 10 14 15 18 10 15
Evenness AA 0.90 0.77 0.73 0.79 NA NA NA NA 0.75 NA NA NA
BD 0.58 0.57 0.75 0.61 0.52 0.52 0.58 0.53 0.53 0.53 0.55 0.69
DM 0.65 0.60 0.84 0.71 0.68 NA 0.65 0.62 0.58 0.60 0.71 0.68
Simpson AA 0.96 0.93 0.92 0.89 0.00 0.00 0.00 0.00 0.34 0.00 0.00 0.00
BD 0.83 0.74 0.91 0.76 0.72 0.70 0.74 0.70 0.71 0.73 0.72 0.81
DM 0.86 0.79 0.93 0.83 0.74 0.00 0.73 0.76 0.75 0.77 0.76 0.79
InvSimpson AA 27.83 13.82 12.67 9.51 1.00 1.00 1.00 1.00 1.51 1.00 1.00 1.00
BD 5.79 3.82 11.65 4.23 3.57 3.38 3.79 3.37 3.42 3.72 3.52 5.17
DM 7.38 4.69 13.99 5.74 3.83 1.00 3.76 4.19 4.01 4.28 4.19 4.69
Fungal communities’ alpha diversity
Indices Locations Fermentation days
0 2 4 6 8 10 12 14 16 18 20 22
Chao1 AA 3 3 3 1 2 1 1 1 1 1 1 1
BD 11 4 2 5 1 1 1 1 1 1 1 1
DM 4 3 2 1 1 1 1 1 1 1 1 1
Evenness AA 0.00 0.09 0.06 NA 0 NA NA NA NA NA NA NA
BD 0.17 0.00 0.01 0.01 NA NA NA NA NA NA NA NA
DM 0.09 0.05 0.08 NA NA NA NA NA NA NA NA NA
Simpson AA 0.00 0.04 0.02 0.00 0 0 0 0 0 0 0 0
BD 0.15 0.00 0.00 0.00 0 0 0 0 0 0 0 0
DM 0.04 0.02 0.02 0.00 0 0 0 0 0 0 0 0
InvSimpson AA 1.00 1.04 1.02 1.00 1 1 1 1 1 1 1 1
BD 1.18 1.00 1.00 1.00 1 1 1 1 1 1 1 1
DM 1.05 1.02 1.02 1.00 1 1 1 1 1 1 1 1

NA (not analyzed) - for the samples exclusively dominated by the one species.

Fig. 5. The shannon index for A) bacterial

and B) fungal communities during the Tej
fermentation period. Each dot in the line
graph represents the Shannon index value
for Tej samples from Addis Ababa (Green),
Bahir Dar (Orange), and Debre Markos (Vi­
olet) households. Principal component
analysis of Unweighted Unifrac distances
demostrated β-divesity for C) bacterial and
D) fungal communities during fermentation
periods. The effects of fermentation days are
demostrated by gradiant color increases.
(For interpretation of the references to color
in this figure legend, the reader is referred to
the Web version of this article.)

4. Conclusions modernization of any spontaneous fermentation system. This research

elucidated important information on the interaction of physicochemical
Fermentation dynamics studies have a vital role for the and microbial properties during Tej fermentation period; this will

8
E.G. Fentie et al. LWT 155 (2022) 112927

contribute a lot to future modernization processes. Most importantly, Bolyen, E., Rideout, J. R., Dillon, M. R., Bokulich, N. A., Abnet, C. C., Al-Ghalith, G. A.,
Alexander, H., Alm, E. J., Arumugam, M., Asnicar, F., Bai, Y., Bisanz, J. E.,
irrespective of sample collection households, all time-course study
Bittinger, K., Brejnrod, A., Brislawn, C. J., Brown, C. T., Callahan, B. J., Caraballo-
samples were dominated by the Lactobacillus and Saccharomyces genera, Rodríguez, A. M., Chase, J., & Caporaso, J. G. (2019). Reproducible, interactive,
especially after day 4 of fermentation. In line with these dominant scalable and extensible microbiome data science using QIIME 2. Nature
fermentative microbes, physicochemical characteristics also agreed with Biotechnology, 37, 852–857. https://doi.org/10.1038/s41587-019-0209-9
Borneman, J., & Hartin, R. J. (2000). PCR primers that amplify fungal rRNA genes from
the findings of previous reports on Tej. Thus, both genera might seri­ environmental samples. Applied and Environmental Microbiology, 66, 4356–4360.
ously considered as potential mixed starter culture for the fermentation https://doi.org/10.1128/AEM.66.10.4356-4360.2000
of Ethiopian honey wine, Tej, without sacrificing any of its unique Buglass, A. J. (2010). Handbook of alcoholic beverages: Technical, analytical and nutritional
aspects, 1. John Wiley & Sons, Ltd.
characteristics. Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., &
Holmes, S. P. (2016). DADA2: High-resolution sample inference from Illumina
CRediT authorship contribution statement amplicon data. Nature Methods, 13, 581–583. https://doi.org/10.1038/nmeth.3869
* Cason, E. D., Mahlomaholo, B. J., Taole, M. M., Abong, G. O., Vermeulen, J. G., de
Smidt, O., Vermeulen, M., Steyn, L., Valverde, A., & Viljoen, B. (2020). Bacterial and
Eskindir Getachew Fentie: Conceptualization, Methodology, Data fungal dynamics during the fermentation process of Sesotho, a traditional beer of
curation, Formal analysis, Software, Visualization, Writing – original Southern Africa. Frontiers in Microbiology, 11, 1451. https://doi.
org/10.3389/fmicb.2020.01451.
draft. Minsoo Jeong: Data curation, Software, Visualization. Shimelis Cha, I.-T., Lee, H.-W., Song, H. S., Yim, K. J., Kim, K.-N., Kim, D., Roh, S. W., & Nam, Y.-
Admassu Emire: Conceptualization, Writing – review & editing, Su­ D. (2014). Diversity of lactic acid bacteria in the Korean traditional fermented
pervision. Hundessa Dessalegn Demsash: Conceptualization, Writing beverage shindari, determined using a culture-dependent method. Current Topic in
Lactic Acid Bacteria and Probiotics, 2, 34–37. https://doi.org/10.35732/
– review & editing, Supervision. Min A. Kim: Data curation. Jae-Ho ctlabp.2014.2.1.34
Shin: Conceptualization, Writing – review & editing, Resources, Chen, G. L., Munyao Mutie, F., Xu, Y. B., Saleri, F. D., Hu, G. W., & Guo, M. Q. (2020).
Supervision. Antioxidant, anti-inflammatory activities and polyphenol profile of Rhamnus
prinoides. Pharmaceuticals, 13, 55. https://doi.org/10.3390/ph13040055
Chidi, B. S., Bauer, F. F., & Rossouw, D. (2018). Organic acid metabolism and the impact
Declaration of competing interest of fermentation practices on wine acidity: A review. South African Journal for Enology
& Viticulture, 39. https://doi.org/10.21548/39-2-3172
Chua, L. S., Rahaman, N. L. A., Adnan, N. A., & Eddie Tan, T. T. (2013). Antioxidant
The authors declare that they have no known competing financial activity of three honey samples in relation with their biochemical components.
interests or personal relationships that could have appeared to influence Journal of Analytical Methods in Chemistry, 313798. https://doi.org/10.1155/2013/
the work reported in this paper. 313798
Cristian, H., Carlos, J., & Marta, Q. (2015). Evaluation physicochemical and sensory
aspects of mead, produced by different nitrogen sources and commercial yeast.
Acknowledgments Chemical Engineering Transactions, 43, 1–6. https://doi.org/10.3303/CET1543001
Di Maro, E., Ercolini, D., & Coppola, S. (2007). Yeast dynamics during spontaneous wine
fermentation of the Catalanesca grape. International Journal of Food Microbiology,
Eskindir Getachew Fentie would like to acknowledge Addis Ababa 117, 201–210. https://doi.org/10.1016/j.ijfoodmicro.2007.04.007
University, Addis Ababa Science and Technology University, and * Diaz, M., Kellingray, L., Akinyemi, N., Adeiranye, O. O., Olaonipekun, A., Bayili, G.,
Kyungpook National University. This work was supported by the Com­ Ibezim, J., Salomina du plessis, A., Houngbédji, M., Kamya, D., Mukisa, I., Mulaw,
G., Josiah, S., Chienjo, W., Atter, A., Agbemale, E., Annan, T., Ackah, B., & Buys, E.
mercializations Promotion Agency for R&D Outcomes (COMPA) grant (2019). Comparison of the microbial composition of African fermented foods using
funded by the Ministry of Science and ICT (R&D project No. amplicon sequencing. Scientific Reports, 9, 13863. https://doi.
1711139487), the project to train professional personnel in biological org/10.1038/s41598-019-50190-4.
Fentie, E. G., Emire, S. A., Demsash, H. D., Dadi, D. W., & Shin, J.-H. (2020). Cereal- and
materials by the Ministry of Environment, and Korea Basic Science fruit-based Ethiopian traditional fermented alcoholic beverages. Foods, 9, 1781.
Institute (National research Facilities and Equipment center) grand https://doi.org/10.3390/foods9121781
funded by the Ministry of Education (2021R1A6C101A416), South Fuochi, V., Coniglio, M. A., Laghi, L., Rescifina, A., Caruso, M., Stivala, A., &
Furneri, P. M. (2019). Metabolic characterization of supernatants produced by
Korea. Lactobacillus spp. with in vitro anti-Legionella activity. Frontiers in Microbiology, 10,
1403. https://doi.org/10.3389/fmicb.2019.01403
References Gardes, M., & Bruns, T. D. (1993). ITS primers with enhanced specificity for
basidiomycetes—application to the identification of mycorrhizae and rusts.
Molecular Ecology, 2, 113–118. https://doi.org/10.1111/j.1365-294X.1993.tb00005.
Adamenko, K., Kawa-Rygielska, J., Kucharska, A. Z., & Piórecki, N. (2018).
x
Characteristics of biologically active compounds in cornelian cherry meads.
Gebremedhin, G., Tadesse, G., & Kebede, E. (2013). Physiochemical characteristics of
Molecules, 23, 2024. https://doi.org/10.3390/molecules23082024
honey obtained from traditional and modern hive production systems in Tigray
Akalın, H., Bayram, M., & Anlı, R. E. (2017). Determination of some individual phenolic
region, northern Ethiopia. Momona Ethiopian Journal of Science, 5, 115–125. https://
compounds and antioxidant capacity of mead produced from different types of
doi.org/10.4314/mejs.v5i1.85335
honey: Properties of mead from different types of honey. Journal of the Institute of
Haig, S. J., Quince, C., Davies, R. L., Dorea, C. C., & Collins, G. (2015). The relationship
Brewing, 123, 167–174. https://doi.org/10.1002/jib.396
between microbial community evenness and function in slow sand filters. mBio, 6.
Aranda, Orozco, P., & Matallana. (2019). Yeast life span and its impact on food
https://doi.org/10.1128/mBio.00729-15. e00729–e00715.
fermentations. Fermentatio, 5, 37. https://doi.org/10.3390/fermentation5020037
He, M., Wu, B., Qin, H., Ruan, Z., Tan, F., Wang, J., Shui, Z., Dai, L., Zhu, Q., Pan, K.,
Arnao, M. B. (2000). Some methodological problems in the determination of antioxidant
Tang, X., Wang, W., & Hu, Q. (2014). Zymomonas mobilis: A novel platform for
activity using chromogen radicals: A practical case. Trends in Food Science &
future biorefineries. Biotechnology for Biofuels, 7, 1–15. https://doi.org/10.1186/
Technology, 11, 419–421. https://doi.org/10.1016/S0924-2244(01)00027-9
1754-6834-7-101
* Bahiru, B., Mehari, T., & Ashenafi, M. (2001). Chemical and nutritional properties of
Holzapfel, W. (2015). Advances in fermented foods and beverages; improving quality,
‘tej’, an indigenous Ethiopian honey wine: Variations within and between
technologies and health benefits (1st ed.). Woodhead Publishing.
production units. Journal of Food Technology in Africa, 6, 104–108. https://doi.
Iglesias, A., Pascoal, A., Choupina, A. B., Carvalho, C. A., Feás, X., & Estevinho, L. M.
org/10.4314/jfta.v6i3.19299.
(2014). Developments in the fermentation process and quality improvement
* Bahiru, B., Mehari, T., & Ashenafi, M. (2006). Yeast and lactic acid flora of tej, an
strategies for mead production. Molecules, 19, 12577–12590. https://doi.org/
indigenous Ethiopian honey wine: Variations within and between production units.
10.3390/molecules190812577
Food Microbiology, 23, 277–282. https://doi.org/10.1016/j.fm.2005.05.007.
Jaishankar, J., & Srivastava, P. (2017). Molecular basis of stationary phase survival and
Balogu, T., & Towobola, O. (2017). Production and quality analysis of wine from honey
applications. Frontiers in Microbiology, 8. https://doi.org/10.3389/
and coconut milk blend using Saccharomyces cerevisiae. Fermentatio, 3, 16. https://
fmicb.2017.02000, 2000.
doi.org/10.3390/fermentation3020016
Jung, M. J., Nam, Y. D., Roh, S. W., & Bae, J. W. (2012). Unexpected convergence of
Bartowsky, E. J., Xia, D., Gibson, R. L., Fleet, G. H., & Henschke, P. A. (2003). Spoilage of
fungal and bacterial communities during fermentation of traditional Korean
bottled red wine by acetic acid bacteria. Letters in Applied Microbiology, 36, 307–314.
alcoholic beverages inoculated with various natural starters. Food Microbiology, 30,
https://doi.org/10.1046/j.1472-765X.2003.01314.x
112–123. https://doi.org/10.1016/j.fm.2011.09.008
Belay, A., Haki, G. D., Birringer, M., Borck, H., Lee, Y.-C., Cho, C.-W., Kim, K.-T.,
Jung, Y., Tagele, S. B., Son, H., Ibal, J. C., Kerfahi, D., Yun, H., Lee, B., Park, C. Y.,
Bayissa, B., Baye, K., & Melaku, S. (2017). Sugar profile and physicochemical
Kim, E. S., Kim, S. J., & Shin, J. H. (2020). Modulation of gut microbiota in Korean
properties of Ethiopian monofloral honey. International Journal of Food Properties, 20,
navy trainees following a healthy lifestyle change. Microorganisms, 8, 16. htt
2855–2866. https://doi.org/10.1080/10942912.2016.1255898
p://doi:10.3390/microorganisms8091265w.
Blois, M. S. (1958). Antioxidant determinations by the use of a stable free radical. Nature,
181, 1199–1200. https://doi.org/10.1038/1811199a0

9
E.G. Fentie et al. LWT 155 (2022) 112927

Kim, D.-H., Chon, J.-W., Kim, H., Kim, H.-S., Choi, D., Hwang, D.-G., & Seo, K.-H. (2015). * Ramos, C. L., de Almeida, E. G., Pereira, G. V. de M., Cardoso, P. G., Dias, E. S., &
Detection and enumeration of lactic acid bacteria, acetic acid bacteria and yeast in Schwan, R. F. (2010). Determination of dynamic characteristics of microbiota in a
kefir grain and milk using quantitative real-time PCR: Quantitative PCR for kefir fermented beverage produced by Brazilian Amerindians using culture-dependent
microorganism. Journal of Food Safety, 35, 102–107. https://doi.org/10.1111/ and culture-independent methods. International Journal of Food Microbiology, 140,
jfs.12153 225–231. https://doi.org/10.1016/j.ijfoodmicro.2010.03.029.
Lane, D. L. (1991). 16S/23S rRNA sequencing. In E. Stackebrandt, & M. Goodfellow Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C. (1999).
(Eds.), Nucleic acid techniques in bacterial systematics (pp. 115–175). John Wiley & Antioxidant activity applying an improved ABTS radical cation decolorization assay.
Sons. Free Radical Biology and Medicine, 26, 1231–1237. https://doi.org/10.1016/S0891-
Lemi, B. W. (2020). Microbiology of Ethiopian traditionally fermented beverages and 5849(98)00315-3
condiments. International Journal of Microbiology, 1–8. https://doi.org/10.1155/ Sha, S. P., Suryavanshi, M. V., & Tamang, J. P. (2019). Mycobiome diversity in
2020/1478536 traditionally prepared starters for alcoholic beverages in India by high-throughput
Marshall, E., & Mejía-Loríio, D. J. (Eds.). (2011). Traditional fermented food and beverages sequencing method. Frontiers in Microbiology, 10, 348. https://doi.org/10.3389/
for improved livelihoods. Rural Infrastructure and Agro-Industries Division. Food and fmicb.2019.00348
Agriculture Organization of the United Nations. Steinkraus, K. H. (Ed.). (2004). Industrialization of indigenous fermented foods (2nd ed.).
Martínez-Moreno, R., Morales, P., Gonzalez, R., Mas, A., & Beltran, G. (2012). Biomass Marcel Dekker. rev.expanded.
production and alcoholic fermentation performance of Saccharomyces cerevisiae as Tamang, J. P., & Thapa, S. (2006). Fermentation dynamics during production of bhaati
a function of nitrogen source. FEMS Yeast Research, 12, 477–485. https://doi.org/ jaanr, a traditional fermented rice beverage of the Eastern Himalayas. Food
10.1111/j.1567-1364.2012.00802.x Biotechnology, 20, 251–261. https://doi.org/10.1080/08905430600904476
Ma, C., Sun, Z., Chen, C., Zhang, L., & Zhu, S. (2014). Simultaneous separation and Taylor, T. N., Krings, M., & Taylor, E. L. (2015). Fossil fungi. Academic Pr.
determination of fructose, sorbitol, glucose and sucrose in fruits by HPLC–ELSD. Tôrres, A. R., da Silva Lyra, W., de Andrade, S. I. E., Andrade, R. A. N., da Silva, E. C.,
Food Chemistry, 145, 784–788. https://doi.org/10.1016/j.foodchem.2013.08.135 Araújo, M. C. U., & da Nóbrega Gaião, E. (2011). A digital image-based method for
Motlhanka, K., Zhou, N., & Lebani, K. (2018). Microbial and chemical diversity of determining of total acidity in red wines using acid–base titration without indicator.
traditional non-cereal based alcoholic beverages of sub- Saharan Africa. Beverages, 4, Talanta, 84, 601–606. https://doi.org/10.1016/j.talanta.2011.02.002
36. https://doi.org/10.3390/beverages4020036 Turenne, C. Y., Sanche, S. E., Hoban, D. J., Karlowsky, J. A., & Kabani, A. M. (1999).
Mulaw, G., & Tesfaye, A. (2017). Technology and microbiology of traditionally Rapid identification of fungi by using the ITS2 genetic region and an automated
fermented food and beverage products of Ethiopia: A review. African Journal of fluorescent capillary electrophoresis system. Journal of Clinical Microbiology, 37,
Microbiology Research, 11, 825–844. https://doi.org/10.5897/AJMR2017.8524 1846–1851. https://doi.org/10.1128/JCM.37.6.1846-1851.1999
Muyzer, G., de Waal, E. C., & Uitterlinden, A. G. (1993). Profiling of complex microbial Vogel, & Gobezie. (1995). Indigenous fermented foods in which ethanol is a major
populations by denaturing gradient gel electrophoresis analysis of polymerase chain product. In Hand book of Indigenous fermented foods (2nd ed., pp. 365–373). Marcel
reaction-amplified genes coding for 16S rRNA. Applied and Environmental Dekker Inc.
Microbiology, 59, 695–700. https://doi.org/10.1128/aem.59.3.695-700.1993 Walker, G., & Stewart, G. (2016). Saccharomyces cerevisiae in the production of
* Nemo, R., & Bacha, K. (2020). Microbial, physicochemical and proximate analysis of fermented beverages. Beverages, 2, 30. https://doi.org/10.3390/beverages2040030
selected Ethiopian traditional fermented beverages. Lebensmittel-Wissenschaft und Weckx, S., Van der Meulen, R., Maes, D., Scheirlinck, I., Huys, G., Vandamme, P., & De
-Technologie- Food Science and Technology, 131. https://doi.org/10.1016/j. Vuyst, L. (2010). Lactic acid bacteria community dynamics and metabolite
lwt.2020.109713. production of rye sourdough fermentations share characteristics of wheat and spelt
* Nemo, R., & Bacha, K. (2021). Microbial dynamic and growth potential of selected sourdough fermentations. Food Microbiology, 27, 1000–1008. https://doi.org/
pathogens in Ethiopian traditional fermented beverages. Annals of Microbiology, 71, 10.1016/j.fm.2010.06.005
22. https://doi.org/10.1186/s13213-021-01635-7. This report focusses on bacterial and fungal dynamics of South African’s traditional beer.
OIV. (2020). Compendium of international methods of wine and must analysis. OVI. It gave us a very important information on how to tackle the microbial dynamics
Oksanen, J., Kindt, R., Legendre, P., O’Hara, B., Stevens, M. H. H., Oksanen, M., & segment our research
Suggests, M. (2017). The vegan package. Community Ecological Package, 10, 719. This report was the first report to analyze Tej using culture independent microbial
Parada, A. E., Needham, D. M., & Fuhrman, J. A. (2016). Every base matters: Assessing characterization method. Even though it was entirely focused on the final product
small subunit rRNA primers for marine microbiomes with mock communities, time microbial characterization, it was still a good input for our study.
series and global field samples: Primers for marine microbiome studies. This report deals on physicochemical and microbiological characterization of selected
Environmental Microbiology, 18, 1403–1414. https://doi.org/10.1111/1462- Ethiopia traditional beverages. Tej was one of the beverages included in this study;
2920.13023 Thus, we used some of the information revealed on this report as a springboard for
Park, E. J., Chang, H. W., Kim, K. H., Nam, Y. D., Roh, S. W., & Bae, J. W. (2009). our study.
Application of quantitative real-time PCR for enumeration of total bacterial, This report gives information on the fermentative and pathogenic microbial dynamics
archaeal, and yeast populations in kimchi. Journal of Microbiology, 47, 682–685. using culture dependent methods for selected Ethiopian traditional beverages. Since
https://doi.org/10.1007/s12275-009-0297-1 Tej was one of them, we used it as a good source of information for further analysis
Pereira, A. P., Oliveira, J. M., Mendes-Ferreira, A., Estevinho, L. M., & Mendes-Faia, A. our results.
(2017). Mead and other fermented beverages. In Current developments in This report focuses on the microbial dynamics of Cauim (kawi). It gave us a very
biotechnology and bioengineering (pp. 407–434). Elsevier. https://doi.org/10.1016/ important insight on the importance of quantitative and qualitative microbial study
B978-0-444-63666-9.00014-5. on the dynamics of any fermentation system
Plavec, T. V., & Berlec, A. (2020). Safety aspects of genetically modified lactic acid
bacteria. Microorganisms, 8. https://doi.org/10.3390/microorganisms8020297

10