Bioresource Technology Reports 17 (2022) 100891

Contents lists available at ScienceDirect

Bioresource Technology Reports

journal homepage: www.sciencedirect.com/journal/bioresource-technology-reports

Treatment of dairy industry wastewater using bacterial biomass isolated

from eutrophic lake sediments for the production of agricultural water
María Custodio *, Richard Peñaloza , Ciro Espinoza , Wilson Espinoza , Juana Mezarina
Universidad Nacional del Centro del Perú, Av. Mariscal Castilla N◦ 3909, Huancayo, Peru.

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

Keywords: In the present study, the treatment of dairy industry wastewater using bacterial biomass isolated from eutrophic
Microbial biomass lake sediments was carried out to produce water for agricultural use. Microorganisms were isolated from lake
Wastewater sediment to establish bacterial consortiums efficient in the removal of organic matter from dairy wastewater. The
Dairy industry
nature of the effluents varied according to the treatment system and parameter evaluated. In the anaerobic
Aerobic and anaerobic treatment
treatment, chlorides, sulfates, phosphates and chemical oxygen demand recorded the highest average values. In
the aerobic treatment, turbidity, electrical conductivity, ammonium, total suspended solids and biochemical
oxygen demand recorded the highest average values. The analysis of the percentage of similarity of the anaerobic
(BSI1) and aerobic (BSI2) bacterial consortia with a contribution of more than 1% revealed a 91.38% contri­
bution. In anaerobic reactors the orders Clostridiales (14.27%) and Lactobacillales (12.44%) predominated and
in aerobic reactors the order Burkholderiales (20.67%).

1. Introduction dairy products and raw water, pose a potential contamination problem
(Karadag et al., 2015). Frequently used inorganic compounds include
Industries that generate large volumes of wastewater and the release phosphates (used as deflocculants and emulsifiers in cleaning com­
of toxic substances into the environment include the food, distillery and pounds), chlorine (used in detergents and disinfectant products) and
paper industries, among others (Finnegan et al., 2018; Ahmad et al., nitrogen (contained in wetting agents and disinfectants). These condi­
2019). The dairy industry, has a significant impact on the environment tions cause the eutrophication process to increase (Ahmad et al., 2019)
due to excessive water consumption (from 8.0 to 35.0 L of water for kg of and the risk to human health and ecosystems (Stanchev et al., 2020)
milk) and high effluent production (from 0.2 to 10 L of wastewater per Therefore, treatment of dairy wastewater becomes very important to
liter of milk processed) (Mansoorian et al., 2016), revealing the gener­ achieve clean production.
ation of 0.2 to 10 L of wastewater per liter of milk processed (Akansha Studies on wastewater treatment show that it can be a very complex
et al., 2020). The main processes that generate polluting waste are the process due to the wide variety and concentration levels of pollutants
production of cheese, cream and butter and the washing of drying (Katuri et al., 2012; Silva et al., 2019; Munoz-Cupa et al., 2021). In
equipment. The wastewater generated by the dairy industry is generally general, dairy wastewater is treated by physicochemical and biological
neutral or slightly alkaline, with a tendency to become acidic due to methods. Treatment of dairy wastewater by physicochemical methods
lactose fermentation (Shi et al., 2021). generally requires a lot of energy and a large investment in operating
Final wastewater from the dairy industry is characterized by a high costs. Whereas treatment by biological methods is economical, envi­
content of dissolved organic substances such as lactose, mineral salts and ronmentally friendly, highly efficient and the application is not
colloidal protein suspensions, with lactose being the most polluting by- complicated (Charalambous et al., 2020). However, sludge formation
product (Charalambous et al., 2020). The COD in this type of wastewater during biological treatment can represent a disposal problem and raise
ranges between 2000 and 4000 mg/L and a BOD between 2000 and the costs of the process (Ahmad et al., 2019). Other technologies to
3000 mg/L (Herrera and Corpas, 2013; Parra-Huertas and Campos- improve pollutant degradation include the use of microbial fuel cells
Montiel, 2014). In addition, non-edible materials used in the process (Molognoni et al., 2018), where the microorganisms interact with
contain inorganic substances that, by themselves, or added to those in electrodes that use electrons (Negassa et al., 2021) and bioaugmentation

* Corresponding author.
E-mail address: mcustodio@uncp.edu.pe (M. Custodio).

https://doi.org/10.1016/j.biteb.2021.100891
Received 1 October 2021; Received in revised form 16 November 2021; Accepted 18 November 2021
Available online 23 November 2021
2589-014X/© 2021 The Authors. 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/).
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

technology (Lee et al., 2020). was added into a 250 mL Erlenmeyer flask containing 99 mL of sterile
Bioaugmentation utilizes microorganisms that may be indigenous to distilled water and placed on a magnetic stirrer at 150 rpm for 1 h at
a contaminated area or may be isolated from other locations with the room temperature. It was allowed to stand for 10 min and then the su­
ability to remove contaminants (Porwal et al., 2015). The wide range of pernatant was separated. The isolation of aerobic microorganisms was
enzymes produced by various bacteria can enhance the degradation carried out by successive dilutions, enrichment and biochemical
efficiency of target contaminants (Sharma and Shukla, 2021). However, methods. Dilutions from 10− 1 to 10− 5 were prepared by transferring
bioaugmentation is often affected by the type of pollutant, characteris­ 1 mL of the bacterial suspension to 9 mL of sterile distilled water, mixing
tics of the microorganism, and environmental factors. The microor­ and from this dilution; the next dilution was prepared in the same
ganisms selected to formulate an inoculum must not only be able to proportions and so on until the last dilution (10− 5). Seeding of each
degrade specific contaminants but also be competitive and persist after dilution was performed on nutrient agar by surface streaking in dupli­
inoculation. The interaction of microorganisms in a consortium will cate and incubation was performed at 30 ◦ C for 48 h. The selection of the
facilitate their successful establishment and functioning in the envi­ microorganisms was made on the basis of colony morphology. Sub­
ronment, where they will be applied in the bioremediation process. In cultures of each colony were performed on nutrient agar to obtain the
this context and considering that, wastewater treatment has the poten­ target bacterial strain. Isolation of strict anaerobic microorganisms was
tial to become a sustainable process if adequate technologies are adop­ performed from 10 mL of the supernatant in 500 mL of anaerobic culture
ted. The objective of this study was to treat dairy industry wastewater medium according to the protocol of Bond and Lovley (2003). The
using bacterial biomass isolated from eutrophic lake sediments to pro­ media were incubated at 30 ◦ C for 96 h under anaerobic conditions. The
duce water for agricultural use. media were incubated at 30 ◦ C for 96 h under anaerobic conditions in a
Thermo Fisher CO2 cell culture incubator (Madrid, España).
2. Materials and methods
2.5. Treatment of dairy wastewater using bacterial consortiums
2.1. Collection of dairy wastewater samples
2.5.1. Preparation of inoculum
Individual samples of dairy wastewater were collected in plastic From the isolated aerobic bacterial strains, inocula were prepared by
bottles (3 L) from dairy processing plants in the Usibamba transferring the active culture strain (touching each strain with the
(443,400.67 m E - 8,670,494.81 m S, zone 18 L) and Chaquicocha bacteriological loop) to a Schott flask containing 100 mL of brain heart
(444,429.17 m E - 8,669,514.34 m S, zone 18 L).of the district of San infusion broth. Each inoculum was incubated for 24 h at 30 ◦ C and
José de Quero, province of Concepción in the central region of Peru. 200 rpm on orbital shaker with incubation, Professional 3500. After the
Previously, the bottles were washed with alcohol, rinsed with distilled incubation period, the inoculum was diluted with sterile distilled water
water and dried in an oven at 50 ◦ C for 30 min. At the sample collection and spectrophotometric titration was performed at 0.1 at a wavelength
site, the bottles were rinsed with dairy wastewater several times and of 600 nm corresponding to an approximate bacterial concentration of
then filled. The preservation and transport of the dairy wastewater 1 × 106 bacteria/mL. The inoculum of the aerobic bacterial consortium
samples to the laboratory was carried out according to the standard represented by sedimentary bacterial isolates (BSI2) consisted of the
method (APHA et al., 1999). At the Water Research Laboratory of the mixture of each of the strains in equal proportions. The inoculum of the
National University of Central Peru, individual samples were pooled to anaerobic bacterial consortium represented by sedimentary bacterial
form composite samples and stored at 4 ◦ C for analysis. isolates (BSI1) was also diluted to obtain a similar bacterial concentra­
tion as the consortium under aerobic conditions.
2.2. Physicochemical analysis of dairy wastewater
2.5.2. Aerobic treatment
The analytical determinations of the physicochemical parameters The aerobic treatment of dairy wastewater using bacterial consortia
were carried out using standard APHA/AWWA/WEF methods (three was carried out at laboratory scale in 3 L, 12 cm diameter and 55 cm
replicates per parameter). The pH, turbidity and EC were determined height reactors (Fig. 1). The process started with the sterilization of 2 L
using a portable Hanna Instrument (HI 991301 Microprocessor) (Padua, of the wastewater in an autoclave at 121 ◦ C for 15 min, allowed to cool
Italia). Chlorides (Cl− ), phosphate (PO₄3− ), sulfates (SO₄2− ) and to room temperature and incorporated into the respective reactor. Then,
ammonium (NH4+) were determined with the PC-MultiDirect Photom­ 50 mL of inoculum of the aerobic bacterial consortium was added.
eter (Dortmund- Alemania). Previously, the method and range of mea­ Process conditions were pH 7, 20 ◦ C, 150 rpm and an airflow rate of 0.2
surement with cuvette test for a given parameter is selected. Chemical 0.2 L/min in order to maintain the desired oxygen level and achieve
oxygen demand (COD) was determined with the MD 600/MaxiDirect survival, microbial growth and degradation of the contaminant material
photometer (Dortmund- Alemania) and biological oxygen demand (Porwal et al., 2015). The aeration rate was decided based on pre­
(BOD) was determined with the OxiDirect Lovibond (Dortmund- liminary tests. After 72 h, the process was stopped and allowed to stand
Alemania). for 1 h to allow sedimentation of the sludge formed. The effluent was
then removed and filtered to determine the physicochemical
2.3. Collection of lacustrine sediment samples parameters.

Surface sediment samples (10 cm depth) were collected from three 2.5.3. Anaerobic treatment
fish-use ponds in November 2019. Nine sampling stations were estab­ Three glass bottles of 3 L volumetric capacity were used as anaerobic
lished in each lagoon covering the northern, central and southern part. reactors. To each reactor, 2 L of sterile dairy wastewater and 50 mL of
At each sampling station, four sites were selected for sediment sampling. anaerobic microbial consortium were added. Anaerobic conditions were
Sediment samples from each lagoon were collected using a stainless steel ensured by bubbling nitrogen for 5 min in each reactor. The reactors
auger-type device. Samples from each station were mixed to obtain were maintained for 72 h at 20 ◦ C with continuous gentle agitation.
250 g composite samples. Sediment samples were conditioned in sterile Once the operation period had elapsed and the process was completed,
airtight plastic bags and transported on ice to the laboratory for analysis the supernatant was removed and filtered to determine the physico­
chemical parameters of the treated dairy wastewater.
2.4. Isolation of microorganisms from lake sediment

Isolation of microorganisms was performed from 1 g of sediment that

2
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

Fig. 1. Schematic description of the laboratory scale aerobic reactor.

2.6. Molecular characterization of bacterial strains 50 ◦ C for 20 s and extension at 72 ◦ C for 90 s; followed by a final
extension step at 72 ◦ C for 5 min.
2.6.1. DNA extraction The amplification products were visualized on 1.5% agarose gels,
DNA extraction from bacterial isolates was performed using the stained with ethidium bromide as intercalating agent (5 μL) and using
Presto™ Soil DNA Extraction Kit, following the manufacturer’s in­ TAE 1× buffer (Tris, Acetic Acid and EDTA) as migration buffer.
structions and standard protocols. The sample was lysed and the tubes Migration was carried out in an electrophoresis chamber at 90 V and for
were shaken horizontally for 10 min with a vortex. Then, they were an average time of 30 min. Visualization of the bands was performed
centrifuged at 8000 ×g for 2 min in a Sigma model 15 K centrifuge. To using the Transilluminator LED thermoblock. Bacterial 16S rRNA
each tube 150 μL of the second lysis buffer was added, shaken for 5 s and amplicon sequencing of the V3 and V4 hypervariable regions was per­
incubated at 0 ◦ C for 5 min. Then, they were centrifuged at 8000 ×g for formed using the standard Illumina MiSeq v2 next-generation protocol
1 min followed by inhibitor removal and again, centrifuged at 16,000 ×g (ADMERA HEALTH LLC, USA).
for 1 min at room temperature. The liquid that passed through the col­
umn was retained to perform DNA binding to silica membranes. 900 μL 2.6.3. Bioinformatics analysis of sequence reads
of DNA binding buffer was added to the liquid obtained and then 750 μL FASTQ files generated by the FASTQC v0.11.9 program were pro­
was passed onto a DNA binding silica column. Centrifuged at 16,000 ×g cessed for read length, base quality and percentage of nucleotide bases.
for 1 min at room temperature, discarding the liquid that passed through Quality filtering and removal of primer regions and adapters present in
the column. The DNA was washed successively with SL3 buffer and the reads was performed using the Trimmomatic v0.39 program with
Wash buffer, with each wash step centrifuged at 16,000 ×g for 30 s. minimum trimming values of Q30 and trimming of reads less than 30 bp.
Finally, the silica column was centrifuged at 16,000 ×g for 2 min at room All individual reads were greater than 150,000 per isolate with a read
temperature. DNA elution was performed by adding 100 μL of elution length of 251 nucleotides. Taxonomic analysis was performed with the
buffer to the column. Then, it was centrifuged for 2 min at 16,000 ×g at program, based on the minikraken_20171019_4GB database. This pro­
room temperature. The eluted DNA was collected in a 1.5 mL centrifuge gram also handles multiple sequences for circular representation.
microtube. Finally, operational taxonomic units were identified and abundances
were calculated. The 16S rRNA gene sequences reported in this study
2.6.2. PCR amplification of 16S rRNA genes and sequencing were sent to the GenBank database (https://www.ncbi.nlm.nih.gov/bi
PCR amplification was performed using Gene One and GE Healthcare oproject/PRJNA682317/) with the access number PRJNA682317.
Life Sciences kits. 1 μL of universal 16S rRNA F primer, 1 μL of universal
16S rRNA R primer, 22 μL of PCR mix (containing premix buffer, MgCl2,
dNTPs and taqPolymerase) and 1 μL of DNA sample were mixed for a 2.7. Statistical analysis
total reaction volume of 25 μL. Primers 27F (5′ -AGA­
GAGTTTGATCCTGGCTCAG-3′ ) and 1492R (5′ -GGTTACCTTGTTAC­ Similarity percentage analysis (SIMPER) was performed to identify
GACTT-3′ ) were used and amplified for a product of about 1465 bp. PCR the most abundant OTUs for each bacterial isolate, using the relative
was performed on a ThermoScientific SimpliAmp thermal cycler and abundances of the OTUs. The linear discriminant analysis (LDA) effect
consisted of an initial denaturation step at 95 ◦ C for 5 min; followed by size (LEfSe) method was used to identify the most biologically infor­
35 cycles consisting of denaturation at 94 ◦ C for 30 s, hybridization at mative traits that differentiate two or more phenotypes, based on a
normalized relative abundance matrix (Fang et al., 2018). Co-inertia

3
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

analysis was performed to evaluate simultaneous relationships and bacterial consortium are more demanding in the use of ions for their
structural trends in two data sets (bacterial distribution and physico­ growth and survival, characteristic of bacteria of the order Bur­
chemical characteristics of the effluent), optimizing the resulting scores kholderiales. It was also observed that the ions present in the water, such
and relationship model fitting and variable selection (Min et al., 2019) as sulfates, phosphates and ammonium, had a selective decrease in
using the CANOCO V5 platform. Diversity indices (Shannon, Simpson relation to the type of process. In the anaerobic process with Clos­
and Chao) and rarefaction curves comparing shared OTUs among bac­ tridiales dominance, the Clostridium genus had the highest ammonium
terial consortiums (Muriuki et al., 2021) were processed using Paleon­ removal efficiency, with 17.2% compared to 0.9% in the aerobic pro­
tological STatistics (PAST) software. Functional analysis showing the cess. These results are in agreement with other studies that report that
evolution of physicochemical parameters based on the abundance of aerobic treatment of dairy effluents is less efficient in ammonium
each bacterial taxon was examined using a two-way general linear removal due to rapid growth of filamentous biomass and rapid acidifi­
model (GLM). To meet the normality assumptions of this test, the bac­ cation caused by lactose and low water buffering capacity (Carvalho
terial abundance data were transformed into logarithms (Rubin and Leff, et al., 2013; Kolev Slavov, 2017; Ahmad et al., 2019). The highest sulfate
2007). and phosphate removal efficiency occurred in the aerobic process,
revealing that the genus Delftia uses these components of dairy waste­
3. Results and discussion water as nutrients for further development. These observations are
supported by Yarza et al. (2014) which indicate that species of the genus
3.1. Nutrient removal in anaerobic and aerobic treatment systems Delftia show optimal growth at a concentration of sodium dodecyl sul­
fates, present in river water contaminated with detergents.
Anaerobic (BSI1) and aerobic (BSI2) anaerobic and aerobic dairy The removal efficiency of TSS from dairy effluent after the biological
wastewater treatment systems (reactors) were tested to determine the treatment period was 90.3% in anaerobic reactors and 87.7% in aerobic
nutrient removal performance of dairy effluent (Table 1). Each dairy reactors, with TSS concentrations ranging from 80.7 to 89.5 mg/L and
wastewater influent and effluent of the treatment systems was moni­ from 110.2 to 116.6 mg/L in the two types of reactors, respectively. This
tored. The pH of the influent (pH = 6.3) increased slightly after the difference could be due to the type of organic molecules used by the
treatment processes with the BSI1 (pH = 7.0) and BSI2 (pH = 7.1) bacterial consortiums under anaerobic and aerobic conditions. This
consortiums. The results showed that the bacterial consortiums were difference in the removal of TSS is due to the predominance of some
able to reduce the values of turbidity, TSS, EC, BOD and COD. The sig­ bacterial species in each consortium and the preference type of organic
nificant reduction of turbidity and TSS values revealed the influence of compound. Bacteria of the genus Delftia that predominated in aerobic
the sedimentation and filtration processes of the treatment systems. reactors use lactose and fatty acids for their development (Jorgensen
Turbidity reduction in the BSI1 and BSI2 treatment systems was et al., 2009). The orders Lactobacillales with the species Leuconostoc
96.3% (BSI1 treatment) and 95.3% (BSI2 treatment). These results mesenteroides and Clostridia with the species Clostridium tyrobutyricum
reveal the consumption of organic substances and suspended particulate and Clostridium cellulovorans predominated in anaerobic reactors and
matter by microorganisms for growth and survival (Porwal et al., 2015). can degrade refractory organic substances (Fang et al., 2018) and where
These observations are consistent with the results obtained by Rivas the ammonia concentration is high (De Vrieze et al., 2015), explaining
et al. (2011) who, complementing biodegradation with coagulant re­ the effectiveness in ammonium reduction in anaerobic treatment
agents, recorded a turbidity removal of 98% in aerobic processes and systems.
according to the results of Thirugnanasambandham and Ganesamoorthy COD represents the total concentration of organic matter in waste­
(2019) 93% in anaerobic processes. water, and is an important parameter for evaluating the efficiency of
The aerobic process was slightly more efficient in EC removal biological wastewater treatment (Wang et al., 2020; Yang et al., 2021).
(removal efficiency = 68.8%) compared to the anaerobic process The present results revealed that the aerobic treatment had a higher
(removal efficiency = 68.1%). These observations are in agreement with COD removal efficiency (removal efficiency = 80%) than the anaerobic
the results of Wang et al. (2020) who recorded a conductivity removal processes (removal efficiency = 58.6%). These results coincide with
efficiency of 47% in aerobic processes. These results reveal that the BSI2 those reported by Lo and Liao (1989) who recorded in anaerobic

Table 1
Dairy wastewater composition and nutrient removal efficiency by bacterial consortiums in anaerobic and aerobic treatment systems for dairy wastewater.
Parameter Dairy wastewater Anaerobic treatment system Aerobic treatment system
Consortium BSI1 Consortium BSI2

Mean ± SD % Removal Mean ± SD % Removal

pH Influent 6.3 ± 0.2 6.3 ± 0.2

Effluent 7.0 ± 0.2 7.1 ± 0.2
Turbidity (NTU) Influent 946.2 ± 42.8 969.9 ± 42.3
Effluent 34.9 ± 1.4 96.3 45.9 ± 1.9 95.3
EC (μS/cm) Influent 534.9 ± 23.5 551.5 ± 16.9
Effluent 170.7 ± 3.2 68.1 171.9 ± 2.9 68.8
Chlorides (mg/L) Influent 202.5 ± 3.8 197.9 ± 8
Effluent 172 ± 4 15.1 171.5 ± 12.7 13.3
Sulfates (mg/L) Influent 92 ± 1.2 89.4 ± 2.9
Effluent 83.9 ± 1.6 8.8 75.3 ± 2.1 15.8
Phosphates (mg/L) Influent 44.7 ± 1.8 45.4 ± 0.7
Effluent 42.6 ± 0.6 4.7 37.4 ± 0.6 17.6
Ammonium (mg/L) Influent 63.9 ± 1.1 65 ± 2.3
Effluent 52.9 ± 1.5 17.2 64.4 ± 1.2 0.9
TSS (mg/L) Influent 885.6 ± 69.4 929.1 ± 7.7
Effluent 85.8 ± 4.6 90.3 114 ± 3.4 87.7
BOD (mg/L) Influent 1263.75 ± 44.25 1265.1 ± 36.3
Effluent 584.8 ± 23.13 53.7 597.6 ± 12.3 52.8
COD (mg/L) Influent 2011.1 ± 21.1 1995.8 ± 35.9
Effluent 833.2 ± 29.1 58.6 399.5 ± 14.3 80.0

4
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

processes a COD removal efficiency ranging from 46.3 to 62.6% and in contribution. In anaerobic reactors, the predominant orders in the BSI1
aerobic processes a removal efficiency ranging from 93 to 95%. Our consortium were Clostridiales (14.27%) and Lactobacillales (12.44%).
results are also supported by Porwal et al. (2015), who reported COD In aerobic reactors the most representative order was Burkholderiales
removal efficiency ranging from 66.95 to 85.10% in aerobic processes with a 20.67% contribution. Other bacterial orders of interest with
with microbial isolates. Other studies that agree with our observations contributions of less than 7% were representatively similar in both
refer that biodegradation of organic matter in aerobic systems is usually consortiums. Pseudomonadales and Enterobacterales revealed their
faster and more efficient than anaerobic degradation (Biel-Maeso et al., facultative capacity in oxygen use. Pseudomonadales can grow opti­
2019; Azaroff et al., 2021). mally under aerobic conditions. However, some species can live in the
In the present study, bacteria of the order Burkholderiales were presence of nitrate and use nitrogen as an electron acceptor (Palleroni
found to be dominant in aerobic reactors and responsible for higher COD and Brunswick, 2010; Li et al., 2020). They are metabolically versatile
reduction efficiency. These observations reveal the broad enzymatic capable of utilizing a wide range of organic and inorganic compounds
capacity of Burkholderiales to degrade organic substances. The removal (Brzeszcz and Kaszycki, 2018). Caulobacterales, Flavobacteriales,
efficiency of BOD in both wastewater treatment systems was similar, but Sphingobacteriales, Sphingomonadales, and Corynebacteriales pre­
lower than that recorded for COD. Effective reduction in sludge gener­ sented contributions of less than 5% in reactors with BSI2 consortium.
ation and good sedimentation properties have been reported for cheese Bacteroidales and Alteromonadales presented a contribution of less than
whey wastewater precoagulated by aerobic biodegradation (Rivas et al., 4.7% in reactors with BSI1 consortium.
2011). However, in mixed (anaerobic and aerobic) treatment of period
sequences in batch reactors, biological oxygen demand (BOD) removal
efficiencies greater than 90% were observed over a range of operating 3.3. Bacterial diversity in aerobic and anaerobic reactors
temperatures (10–30 ◦ C) (Lo et al., 1986).
In general, the results reveal that treatment of dairy wastewater with Fig. 2 shows the species rarefaction curves as a function of the
bacterial consortiums yields water for agricultural use. The physico­ number of sequences per sample. In both bacterial consortium BSI1 and
chemical parameters of the dairy effluent from the BSI1 and BSI2 BSI2 the rarefaction curves reveal a slight upward trend of richness in
treatment systems are in the range of the U.S. Environmental Protection the anaerobic consortium, but comparisons between diversity indices
Agency’s water reuse guidelines (EPA, 2012). Comparing the quality of were not statistically significant (p > 0.05) (Table 2).
treated dairy wastewater with the maximum permissible effluent limits Fig. 3 shows the bacterial taxa of both consortium, BSI1(anaerobes)
of domestic wastewater treatment plants (MINEN, 2010), it is evident and BSI2 (aerobes), analyzed by Linear discriminant analysis (LDA) ef­
that the effluent quality complies with the regulatory limits of interna­ fect size (LEfSe) which explains where the taxa that best describe a
tional and national regulations. But, if we compare it with the envi­ particular group are found. The change in relative abundance showed
ronmental quality standards for water from rivers and lagoons, class 3 that both bacterial consortiums are statistically significant and are
intended for vegetable irrigation and animal drinking of the Peruvian represented on the horizontal axis. The analysis revealed that the BSI1
regulations (Ministry of Environment, 2017), the average values of BOD and BSI2 consortiums presented 17 and 3 taxa, respectively.
and COD exceed the standards of the receiving water bodies.
Table 2
Comparison of bacterial diversity indices of anaerobic and aerobic treatment
3.2. Composition of bacterial consortiums in aerobic and anaerobic systems.
reactors Index Simpson (Mean ± SD) Shannon Chao
(Mean ± SD) (Mean ± SD)
Fifteen dominant bacterial orders were identified in the anaerobic BSI1 0.728 ± 0.016 1.516 ± 0.068 71.627 ± 19.390
and aerobic dairy wastewater treatment systems. SIMPER analysis of BSI2 0.699 ± 0.111 1.479 ± 0.387 48.693 ± 16.224
anaerobic (BSI1) and aerobic (BSI2) bacterial consortiums with a p Valor 0.663 0.663 0.383
contribution greater than 1% revealed a 91.38% cumulative

Fig. 2. Rarefaction analysis of bacterial 16S rDNA V3 - V4 variable regions from anaerobic and aerobic treatment systems.

5
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

Fig. 3. Linear discriminant analysis (LDA) effect size (LEfSe) indicating enriched bacterial genera associated either with aerobic (green) or anaerobic (red) treatment.
(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4 shows the taxa that were found to be significantly different in identified were Clostridium tyrobutyricum, Clostridium cellulovorans and
the anaerobic and aerobic reactors and the respective hierarchical po­ Clostridium argentinense. In the orders Thermoanabacterales and Veillo­
sition. Visualization of the characteristic taxa of the BSI1 and BSI2 nellales, dominant species have been identified as Thermodesulfobium
consortiums reveals the bacterial composition. The application of Linear narugense and Negativicoccus massiliensis respectively. As well as, some
discriminant analysis (LDA) effect size (LEfSe) identified two significant Proteobacteria with anaerobic tolerance such as Campylobacterales,
differences at phylum level, five at class level and 13 at order level. In Cellvibrionales, Legionallales and Oceanospiralles. Saccharophagus
anaerobic reactors the bacterial consortiums were represented by the degradans belonging to the order Cellvibrionales has been related to
phylum Firmicutes and in aerobic reactors by the phylum Proteobac­ protein hydrolysis and organic matter removal (Fang et al., 2018).
teria. The predominant classes in anaerobic reactors were Negativicutes, Although according to SIMPER analysis its frequency is lower in relation
Clostridia, Bacilli and Epsilonproteobacteria. In aerobic reactors the to the Clostridiales and Bacillales.
class Betaproteobacteria showed higher levels of development with Bacterial diversity was determined based on the orders that
predominance of the order Burkholderiales. The dominant genus was contributed more than 1% to the relative frequency, according to the
Delftia with high oxygen tolerance and ability to utilize mono­ SIMPER analysis. The cumulative percentage of variability explained for
saccharides, disaccharides, amino acids and organic acids (Jorgensen the first two vectors was 87.90, indicating a good fit in the distribution
et al., 2009). In other studies Delftia acidovorans, Delftia tsuruhatensis and of the two matrices used. Fig. 5 shows the sample space of the bacterial
Delftia sp., have shown great ability to metabolize aniline (Arora, 2015), diversity and abundance data (blue arrow) and the projection of the
xenobiotics (Sly et al., 2015) and terephthalate (Shigematsu et al., physicochemical variables of the water (red arrow) to those that are
2003). related to the bacterial order. The length of the arrow indicates the
In anaerobic reactors, the predominant bacterial orders were Lac­ degree of agreement between two sets of data; the shorter the arrow, the
tobacillales, Thermoanaerobacterales, Clostridiales, Selenomonadales lower the consensus between the axes and the parameters studied. The
and Veillonellales. In the order Lactobacillales, the most representative weight of the physicochemical variables of the effluent indicate that the
species was Carnobacterium maltaromaticum, classified as micro­ first axis (X, load of 68.42) determines the differences between the two
aerophilic (Mora et al., 2003). This species is of great interest due to its groups due to the effect of the anaerobic and aerobic treatments.
ability to metabolize organic matter in dairy wastewater, mainly lactose, The variables that experienced significant changes were turbidity,
fat and proteins. In addition, it can inhibit the growth of pathogens due sulfates, phosphates, ammonium, TSS and COD (significant correlation
to its ability to produce bacteriocins (Jacquet et al., 2012). Lactoba­ lower than − 0.7 or higher than 0.7 for the first axis). On the second axis,
cillales convert organic sources into methane through an anaerobic only BOD explains the distribution of bacterial orders. Correlation
process (Najafpour et al., 2009; Karadag et al., 2015), and are essential analysis between water physicochemical variables and bacterial taxa
in power generation and water treatment processes. The role of the whose loadings (RespW) are greater than 40 and rho correlation greater
Clostridia class in the remediation of substrates with high sulfate con­ than 0.4 would explain the derivation of the physicochemical parameter
centrations is important (Gupta and Sar, 2020). Representative species significantly. Of the taxa identified, only two orders were significantly

6
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

Fig. 4. Cladogram plotted from Linear discriminant analysis (LDA) effect size (LEfSe) indicating the phylogenetic distribution of microbiota correlated with the
aerobic (green) and anaerobic (red) treatment. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.)

(p < 0.05) adjusted to the distribution of at least one physicochemical

parameter. The order Burkholderiales with a loading of 51.59 and
dominant in aerobic reactors, correlated positively and significantly
with ammonium concentrations (rho = 0.411), negatively and signifi­
cantly with COD (rho = − 0.406). On the other hand, the order Clos­
tridiales with a load (RespW) of 43.88, dominant in anaerobic reactors,
correlated positively and significantly with phosphate (rho = 0.44). Co-
inertia analysis corroborates that this order significantly influences the
COD removal efficiency of dairy effluents.
The composition of the bacterial communities in the consortium may
vary depending on the conditions specific to the reactor type configu­
ration. The generalized linear model analysis (GLM analysis) shows the
function for the projection of the COD reduction efficiency in relation to
the number of bacteria of the order Burkholderiales. It also indicates that
Burkholderiales and Clostridiales orders are not efficient in ammonium
and phosphate removal, respectively. However, the present results
reveal that mixed bacterial consortiums can help to improve the removal
efficiency of specific parameters (Fig. 6). Bacteria of the orders Fla­
vobacteriales and Bacillales have demonstrated their ability to degrade
phosphates (Zhang et al., 2020) and the orders Bacteroidales and Lac­
tobacillales to degrade ammonium (Gil-Pulido et al., 2018; Ahmad et al.,
2019).

4. Conclusion

Fig. 5. Co-inertia analysis to assess relationships between principal bacterial

In aerobic reactors, the order Burkholderiales showed high perfor­
orders (SIMPER >1% contribution) and dairy effluent parameters treated.
mance in COD removal. In anaerobic reactors, the orders

7
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

Fig. 6. Observed values and Generalized Linear Model (GLM) predictions for mean bacterial density, in order significantly correlated with the concentrations of
physical-chemical parameters of treated effluents (p < 0.05).

Lactobacillales, Thermoanaerobacterales, Clostridiales, Selenomona­ APHA, AWWA, WEF, 1999. Standard Methods for the Examination of Water and
Wastewater, 20th ed. Washington, D. C.
dales and Veillonellales showed better performance in BOD and TSS
Arora, P.K., 2015. Bacterial degradation of monocyclic aromatic amine. Front. Microbiol.
removal. GLM analysis revealed that the composition of bacterial com­ 6, 1–14. https://doi.org/10.3389/fmicb.2015.00820.
munities in the consortiums may vary depending on the conditions Azaroff, A., Monperrus, M., Miossec, C., Gassie, C., Guyoneaud, R., 2021. Microbial
specific to the reactor type configuration. This study reveals that the degradation of hydrophobic emerging contaminants from marine sediment slurries
(Capbreton Canyon) to pure bacterial strain. J. Hazard. Mater. 402, 123477 https://
evaluated bacterial consortiums can be a sustainable and environmen­ doi.org/10.1016/j.jhazmat.2020.123477.
tally friendly technology for wastewater treatment with significant Biel-Maeso, M., González-González, C., Lara-Martín, P.A., Corada-Fernández, C., 2019.
removal of organic pollutants. However, further research is needed to Sorption and degradation of contaminants of emerging concern in soils under
aerobic and anaerobic conditions. Sci. Total Environ. 666, 662–671. https://doi.org/
investigate the behavior of bacterial consortiums in non-sterile dairy 10.1016/j.scitotenv.2019.02.279.
wastewater and under uncontrolled conditions. Bond, D.R., Lovley, D.R., 2003. Electricity production by Geobacter sulfurreducens
attached to electrodes. Appl. Environ. Microbiol. 69, 1548–1555. https://doi.org/
10.1128/AEM.69.3.1548.
CRediT authorship contribution statement Brzeszcz, J., Kaszycki, P., 2018. Aerobic bacteria degrading both n-alkanes and aromatic
hydrocarbons: an undervalued strategy for metabolic diversity and flexibility.
María Custodio: Investigation, Conceptualization, Methodology, Biodegradation 29, 359–407. https://doi.org/10.1007/s10532-018-9837-x.
Carvalho, F., Prazeres, A.R., Rivas, J., 2013. Cheese whey wastewater: characterization
Writing – original draft, Data curation. Richard Peñaloza: Conceptu­ and treatment. Sci. Total Environ. 445–446, 385–396. https://doi.org/10.1016/j.
alization, Methodology, Data curation, Visualization. Ciro Espinoza: scitotenv.2012.12.038.
Investigation, Conceptualization, Methodology, Supervision. Wilson Charalambous, P., Shin, J., Shin, S.G., Vyrides, I., 2020. Anaerobic digestion of industrial
dairy wastewater and cheese whey: performance of internal circulation bioreactor
Espinoza: Investigation, Methodology. Juana Mezarina: Resources, and laboratory batch test at pH 5–6. Renew. Energy 147, 1–10. https://doi.org/
Project administration, Funding acquisition. 10.1016/j.renene.2019.08.091.
De Vrieze, J., Saunders, A.M., He, Y., Fang, J., Nielsen, P.H., Verstraete, W., Boon, N.,
2015. Ammonia and temperature determine potential clustering in the anaerobic
Declaration of competing interest digestion microbiome. Water Res. 75, 312–323. https://doi.org/10.1016/j.
watres.2015.02.025.
EPA, 2012. In: EPA Guidlines for Water Reuse U.S. Environmental Protection Agency.
The authors declare that they have no known competing financial
Guidel. Water Reuse, p. 643.
interests or personal relationships that could have appeared to influence Fang, D., Zhao, G., Xu, X., Zhang, Q., Shen, Q., Fang, Z., Huang, L., Ji, F., 2018. Microbial
the work reported in this paper. community structures and functions of wastewater treatment systems in plateau and
cold regions. Bioresour. Technol. 249, 684–693. https://doi.org/10.1016/j.
biortech.2017.10.063.
Acknowledgment Finnegan, W., Clifford, E., Goggins, J., O’Leary, N., Dobson, A., Rowan, N., Xiao, L.,
Miao, S., Fitzhenry, K., Leonard, P., Tarpey, E., Gil-Pulido, B., Gao, F., Zhan, X.,
This work was funded by CONCYTEC-FONDECYT under the call 2018. DairyWater: Striving for sustainability within the dairy processing industry in
the Republic of Ireland. J. Dairy Res. 85, 366–374. https://doi.org/10.1017/
E041-01 [contract number 76-2018-FONDECYT-BM-IADT-MU]. S0022029918000614.
Gil-Pulido, B., Tarpey, E., Almeida, E.L., Finnegan, W., Zhan, X., Dobson, A.D.W.,
Appendix A. Supplementary data O’Leary, N., 2018. Evaluation of dairy processing wastewater biotreatment in an
IASBR system: aeration rate impacts on performance and microbial ecology.
Biotechnol. Rep. 19, e00263 https://doi.org/10.1016/j.btre.2018.e00263.
Supplementary data to this article can be found online at https://doi. Gupta, A., Sar, P., 2020. Characterization and application of an anaerobic, iron and
org/10.1016/j.biteb.2021.100891. sulfate reducing bacterial culture in enhanced bioremediation of acid mine drainage
impacted soil. J. Environ. Sci. Health A Tox. Hazard. Subst. Environ. Eng. 55,
464–482. https://doi.org/10.1080/10934529.2019.1709362.
References Herrera, O., Corpas, E., 2013. Reduction of contamination in industrial dairy wastewater
using beneficial microorganisms. Biotecnol. Sect. Agropecu. Agroindustrial 11,
Ahmad, T., Aadil, R.M., Ahmed, H., Soares, B.C.V., Souza, S.L.Q., Pimentel, T.C., 57–67.
Scudino, H., Guimarães, J.T., Esmerino, E.A., Freitas, M.Q., Almada, R.B., Jacquet, T., Cailliez-Grimal, C., Francius, G., Borges, F., Imran, M., Duval, J.F.L., Revol-
Vendramel, S.M.R., Silva, M.C., Cruz, A.G., Rahman, U.ur, 2019. Treatment and Junelles, A.M., 2012. Antibacterial activity of class IIa bacteriocin Cbn BM1 depends
utilization of dairy industrial waste: a review. Trends Food Sci. Technol. 88, on the physiological state of the target bacteria. Res. Microbiol. 163, 323–331.
361–372. https://doi.org/10.1016/j.tifs.2019.04.003. https://doi.org/10.1016/j.resmic.2012.04.001.
Akansha, J., Nidheesh, P.V., Gopinath, A., Anupama, K.V., Suresh Kumar, M., 2020. Jorgensen, N.O.G., Brandt, K.K., Nybroe, O., Hansen, M., 2009. Delftia lacustris sp. nov.,
Treatment of dairy industry wastewater by combined aerated electrocoagulation and a peptidoglycandegrading bacterium from fresh water, and emended description of
phytoremediation process. Chemosphere 253. https://doi.org/10.1016/j. Delftia tsuruhatensis as a peptidoglycan-degrading bacterium. Int. J. Syst. Evol.
chemosphere.2020.126652. Microbiol. 59, 2195–2199. https://doi.org/10.1099/ijs.0.008375-0.

8
M. Custodio et al. Bioresource Technology Reports 17 (2022) 100891

Karadag, D., Köroʇlu, O.E., Ozkaya, B., Cakmakci, M., 2015. A review on anaerobic Parra-Huertas, R.A., Campos-Montiel, R.G., 2014. Evaluación del comportamiento de un
biofilm reactors for the treatment of dairy industry wastewater. Process Biochem. 50, reactor UASB con diferentes cargas orgánicas provenientes de lactosuero. Prod. Más
262–271. https://doi.org/10.1016/j.procbio.2014.11.005. Limpia 9, 23–30.
Katuri, K.P., Enright, A.M., O’Flaherty, V., Leech, D., 2012. Microbial analysis of anodic Porwal, H.J., Mane, A.V., Velhal, S.G., 2015. Biodegradation of dairy effluent by using
biofilm in a microbial fuel cell using slaughterhouse wastewater. microbial isolates obtained from activated sludge. Water Resour. Ind. 9, 1–15.
Bioelectrochemistry 87, 164–171. https://doi.org/10.1016/j. https://doi.org/10.1016/j.wri.2014.11.002.
bioelechem.2011.12.002. Rivas, J., Prazeres, A.R., Carvalho, F., 2011. Aerobic biodegradation of precoagulated
Kolev Slavov, A., 2017. Dairy wastewaters – general characteristics and treatment cheese whey wastewater. J. Agric. Food Chem. 59, 2511–2517. https://doi.org/
possibilities – a review. Food Technol. Biotechnol. 55 https://doi.org/10.17113/ 10.1021/jf104252w.
ftb.55.01.17.4520. Rubin, M.A., Leff, L.G., 2007. Nutrients and other abiotic factors affecting bacterial
Lee, J.T.E., Wang, Q., Lim, E.Y., Liu, Z., He, J., Tong, Y.W., 2020. Optimization of communities in an Ohio River (USA). Microb. Ecol. 54, 374–383. https://doi.org/
bioaugmentation of the anaerobic digestion of Axonopus compressus cowgrass for 10.1007/s00248-007-9209-2.
the production of biomethane. J. Clean. Prod. 258, 120932 https://doi.org/ Sharma, B., Shukla, P., 2021. A comparative analysis of heavy metal bioaccumulation
10.1016/j.jclepro.2020.120932. and functional gene annotation towards multiple metal resistant potential by
Li, J., Wu, C., Chen, S., Lu, Q., Shim, H., Huang, X., Jia, C., Wang, S., 2020. Enriching Ochrobactrum intermedium BPS-20 and Ochrobactrum ciceri BPS-26. Bioresour.
indigenous microbial consortiums as a promising strategy for xenobiotics’ cleanup. Technol. 320, 124330 https://doi.org/10.1016/j.biortech.2020.124330.
J. Clean. Prod. 261 https://doi.org/10.1016/j.jclepro.2020.121234. Shi, W., Healy, M.G., Ashekuzzaman, S.M., Daly, K., Leahy, J.J., Fenton, O., 2021. Dairy
Lo, K.V., Liao, P.H., 1989. Anaerobic-aerobic biological treatment of a mixture of cheese processing sludge and co-products: a review of present and future re-use pathways in
whey and dairy manure. Biol. Wastes 28, 91–101. https://doi.org/10.1016/0269- agriculture. J. Clean. Prod. 314, 128035 https://doi.org/10.1016/j.
7483(89)90073-6. jclepro.2021.128035.
Lo, K.V., Chen, W.Y., Liao, P.H., 1986. Mesophilic digestion of screened dairy manure Shigematsu, T., Yumihara, K., Ytaka, U., Numaguchi, M., Morimura, S., Kida, K., 2003.
using anaerobic rotating biological contact reactor. Biomass 9, 81–92. https://doi. Delftia tsuruhatensis sp. nov., a terephthalate assimilating bacterium isolated from
org/10.1016/0144-4565(86)90112-5. activated sludge. Int. J. Syst. Evol. Microbiol. 53, 1479–1483.
Mansoorian, H.J., Mahvi, A.H., Jafari, A.J., Khanjani, N., 2016. Evaluation of dairy Silva, A.N.da, Macêdo, W.V., Sakamoto, I.K., Pereyra, D.de L.A.D., Mendes, C.O.,
industry wastewater treatment and simultaneous bioelectricity generation in a Maintinguer, S.I., Caffaro Filho, R.A., Damianovic, M.H.Z., Varesche, M.B.A.,
catalyst-less and mediator-less membrane microbial fuel cell. J. Saudi Chem. Soc. 20, Amorim, E.L.C.de, 2019. Biohydrogen production from dairy industry wastewater in
88–100. https://doi.org/10.1016/j.jscs.2014.08.002. an anaerobic fluidized-bed reactor. Biomass Bioenergy 120, 257–264. https://doi.
Min, E.J., Safo, S.E., Long, Q., 2019. Penalized co-inertia analysis with applications to- org/10.1016/j.biombioe.2018.11.025.
omics data. Bioinformatics 35, 1018–1025. https://doi.org/10.1093/ Sly, L.I., Wen, A., Fegan, M., 2015. Delftia. In: Bergey’s Manual of Systematics of Archaea
bioinformatics/bty726. and Bacteria, pp. 1–7. https://doi.org/10.1002/9781118960608.gbm00946.
MINEN, 2010. In: Supreme Decree N◦ 003-2010-MINEN. El Peru, p. 11. Stanchev, P., Vasilaki, V., Egas, D., Colon, J., Ponsá, S., Katsou, E., 2020. Multilevel
Ministry of Environment, 2017. Environmental Quality Standards for Water, El Peruano. environmental assessment of the anaerobic treatment of dairy processing effluents in
Molognoni, D., Chiarolla, S., Cecconet, D., Callegari, A., Capodaglio, A.G., 2018. the context of circular economy. J. Clean. Prod. 261 https://doi.org/10.1016/j.
Industrial wastewater treatment with a bioelectrochemical process: assessment of jclepro.2020.121139.
depuration efficiency and energy production. Water Sci. Technol. 77, 134–144. Thirugnanasambandham, K., Ganesamoorthy, R., 2019. Dual treatment of milk
https://doi.org/10.2166/wst.2017.532. processing industry wastewater using electro Fenton process followed by anaerobic
Mora, D., Scarpellini, M., Franzetti, L., Colombo, S., Galli, A., 2003. Reclassification of treatment. Int. J. Chem. React. Eng. 17, 1–10. https://doi.org/10.1515/ijcre-2019-
Lactobacillus maltaromicus (Miller et al. 1974) DSM 20342T and DSM 20344 and 0074.
Carnobacterium piscicola (Collins et al. 1987) DSM 20730T and DSM 20722 as Wang, F., Chen, J., Zhang, C., Gao, B., 2020. Bioresource technology resourceful
Carnobacterium maltaromaticum. Int. J. Syst. Evol. Microbiol. 53, 675–678. https:// treatment of cane sugar industry wastewater by Tribonema minus towards the
doi.org/10.13145/bacdive156165.20201210.5. production of valuable biomass. Bioresour. Technol. 316, 123902 https://doi.org/
Munoz-Cupa, C., Hu, Y., Xu, C., Bassi, A., 2021. An overview of microbial fuel cell usage 10.1016/j.biortech.2020.123902.
in wastewater treatment, resource recovery and energy production. Sci. Total Yang, Y., Hu, Y., Duan, A., Wang, X.C., Hao Ngo, H., Li, Y.Y., 2021. Characterization of
Environ. 754, 142429 https://doi.org/10.1016/j.scitotenv.2020.142429. preconcentrated domestic wastewater toward efficient bioenergy recovery: applying
Muriuki, S.W., Rengan, M.S., Budambula, N.L.M., 2021. Prokaryotic diversity and size fractionation, chemical composition and biomethane potential assay. Bioresour.
potentially pathogenic bacteria in vended foods and environmental samples. Ann. Technol. 319 https://doi.org/10.1016/j.biortech.2020.124144.
Microbiol. 71 https://doi.org/10.1186/s13213-021-01640-w. Yarza, P., Yilmaz, P., Pruesse, E., Glöckner, F.O., Ludwig, W., Schleifer, K.H.,
Najafpour, G.D., Tajallipour, M., Komeili, M., Mohammadi, M., 2009. Kinetic model for Whitman, W.B., Euzéby, J., Amann, R., Rosselló-Móra, R., 2014. Uniting the
an up-flow anaerobic packed bed bioreactor: dairy wastewater treatment. Afr. J. classification of cultured and uncultured bacteria and archaea using 16S rRNA gene
Biotechnol. 8, 3590–3596. https://doi.org/10.5897/AJB2009.000-9355. sequences. Nat. Rev. Microbiol. 12, 635–645. https://doi.org/10.1038/
Negassa, L.W., Mohiuddin, M., Tiruye, G.A., 2021. Treatment of brewery industrial nrmicro3330.
wastewater and generation of sustainable bioelectricity by microbial fuel cell Zhang, M., Pan, L., Liu, L., Su, C., Dou, L., Su, Z., He, Z., 2020. Phosphorus and nitrogen
inoculated with locally isolated microorganisms. J. Water Process Eng. 41, 102018 removal by a novel phosphate-accumulating organism, Arthrobacter sp. HHEP5
https://doi.org/10.1016/j.jwpe.2021.102018. capable of heterotrophic nitrification-aerobic denitrification: Safety assessment,
Palleroni, N.J., Brunswick, N., 2010. Handbook of Hydrocarbon and Lipid Microbiology. removal characterization, mechanism exploration and wastewater treatment.
Handb. Hydrocarb. Lipid Microbiol. https://doi.org/10.1007/978-3-540-77587-4. Bioresour. Technol. 312, 123633 https://doi.org/10.1016/j.biortech.2020.123633.