Chemosphere 262 (2021) 127939

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Optimal influent N-to-P ratio for stable microalgal cultivation in water

treatment and nutrient recovery

gner a, b, *, Clarissa Cazzaniga a, Michael Steidl a, Arnaud Dechesne a,
Dorottya S. Wa

rez a, Benedek Gy Plo
Borja Valverde-Pe
sz a, c, **
a
Department of Environmental Engineering, Technical University of Denmark, Bygningstorvet, Building 115, 2800 Kgs, Lyngby, Denmark
b
Department of Chemistry and Bioscience, Aalborg University, Fredrik Bajers Vej 7H, 9220, Aalborg, Denmark
c
Department of Chemical Engineering, University of Bath, Claverton Down, Bath, BA2 7AY, UK

h i g h l i g h t s

Nitrogen-to-Phosphorus ratio is a powerful tool to control the PBR performance.

Controlling the N-to-P ratio is effective to maintain algal community composition.

Innovative image analysis tool developed to track microbial changes.

a r t i c l e i n f o a b s t r a c t

Article history: Species specific nitrogen-to-phosphorus molar ratio (NPR) has been suggested for green microalgae.
Received 11 June 2020 Algae can store nitrogen and phosphorus, suggesting that the optimum feed concentration dynamically
Received in revised form changes as function of the nutrient storage. We assessed the effect of varying influent NPR on microalgal
27 July 2020
cultivation in terms of microbial community stability, effluent quality and biokinetics. Mixed green
Accepted 5 August 2020
Available online 13 August 2020
microalgae (Chlorella sorokiniana and Scenedesmus sp.) and a monoculture of Chlorella sp. were cultivated
in continuous laboratory-scale reactors treating used water. An innovative image analysis tool, developed
Handling Editor: Y Liu in this study, was used to track microbial community changes. Diatoms proliferated as influent NPR
decreased, and were outcompeted once cultivation conditions were restored to the optimal NPR range.
Keywords: Low NPR operation resulted in decrease in phosphorus removal, biomass concentration and effluent
Nitrogen-to-phosphorus ratio nitrogen concentration. ASM-A kinetic model simulation results agreed well with operational data in the
Algal cultivation absence of diatoms. The failure to predict operational data in the presence of diatoms suggest differences
Algal diversity control in photobioreactors in microbial activity that can significantly influence nutrient recovery in photobioreactors (PBR). No
Process modelling
contamination occurred during Chlorella sp. monoculture cultivation with varying NPRs. Low NPR
Resource recovery
operation resulted in decrease in biomass concentration, effluent nitrogen concentration and nitrogen
Ecological interactions in photobioreactors
quota. The ASM-A model was calibrated for the monoculture and the simulations could predict the
experimental data in continuous operation using a single parameter subset, suggesting stable biokinetics
under the different NPR conditions. Results show that controlling the influent NPR is effective to
maintain the algal community composition in PBR, thereby ensuring effective nutrients uptake.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction

Microalgal cultivation on used water resources has high po-

tential for resource recovery. Cultivation is mostly done in open
* Corresponding author. Department of Environmental Engineering, Technical systems, where the potential of contamination is high. Used water
University of Denmark, Bygningstorvet, Building 115, 2800 Kgs, Lyngby, Denmark. is a potential source of energy, nutrients (nitrogen and phosphorus)
** Corresponding author. Department of Environmental Engineering, Technical and fresh water. However, conventional used water treatment fo-
University of Denmark, Bygningstorvet, Building 115, 2800 Kgs, Lyngby, Denmark. cuses on the destruction of organic and inorganic constituents
E-mail addresses: [email protected] (D.S. W
agner), [email protected]

rez), [email protected] (B.G. Plo

sz).
(Batstone et al., 2015) through energy-intensive processes. Current
(B. Valverde-Pe

https://doi.org/10.1016/j.chemosphere.2020.127939
0045-6535/© 2020 Elsevier Ltd. All rights reserved.
2
gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa

research aims to develop new technologies to reduce energy re- availability, mixing, NPR and inorganic carbon source) can help
quirements (Gao et al., 2014) and facilitate resource recovery maintain the species composition (Borowitzka and Moheimani,
(Verstraete and Vlaeminck, 2011). Microalgae cultivation has been 2013).
proposed as a means to recover resources from used water (Cai The optimal NPR of algae has been an interest since the 1950s,
et al., 2013; Gardner-Dale et al., 2017; Matassa et al., 2015; Mehta when Redfield (1958) found that the optimal NPR in marine
et al., 2015). Combining microalgae cultivation with used water phytoplankton was 16 mol-N/mol-P. Since then, many researchers
resource recovery can result in production of biomass suitable for have suggested that the optimal NPR in microalgae is species spe-
biofuel production (Geiss et al., 2015; Mata et al., 2010; Wa
gner cific (Anbalagan et al., 2016; Beuckels et al., 2015; Rhee and
et al., 2016a). Moreover, microalgal biomass can be used as natu- Gotham, 1980; Whitton et al., 2016) and may vary depending on
ral slow-leaching fertiliser, thereby recycling nutrients present in cultivation conditions, indicating microalgae adaptability to the
the used water (Coppens et al., 2016; Solovchenko et al., 2016), or as culture conditions (Arbib et al., 2013; Beuckels et al., 2015; Boelee
an alternative source of protein (Matassa et al., 2015; Rasouli et al., et al., 2011; Dickinson et al., 2013; Gardner-Dale et al., 2017;
2018). Geider and La Roche, 2002; Liu and Vyverman, 2015; Marcilhac
Many reactor configurations and cultivation methods have been et al., 2015; Rhee, 1978).
proposed to grow microalgae on used water resources. Most of Therefore, the objectives of this study are: (i) to assess the effect
these systems consider open cultivation e.g., (Alca
ntara et al., 2015; of influent NPR on culture composition in a mixed consortium and a
Be
chet et al., 2016; Sforza et al., 2014; Sutherland et al., 2014; Van monoculture during continuous cultivation; (ii) to assess the effect
Den Hende et al., 2014). Variation in the nutrient composition of of algal culture composition on nutrient recovery; (iii) propose PBR
influent water is reported to affect nutrient removal and thus operational strategies to maintain high nutrient recovery efficiency
effluent quality (Arbib et al., 2013). Therefore, using different in open PBR processes.
wastewater streams for microalgae cultivation might compromise
nutrient removal (Wang et al., 2010). Furthermore, nutrient loads 2. Materials and methods
and balances affect algal composition (De Francisci et al., 2018).
When nitrogen or phosphorus are limiting, algae store inorganic 2.1. Microalgae and culture media
carbon as starch and lipids, thus promoting algal biomass into an
appealing feedstock for biofuel production (Ikaran et al., 2015; 2.1.1. Continuous microalgal cultivation using a mixed green
Mayers et al., 2014). Under nutrient limiting conditions, algae also microalgal consortium
shift pigmentation, usually producing more carotenoids, such as A mixed green microalgal consortium, isolated from a natural
lutein, which are valuable antioxidants for food and feed industries pond in contact with wastewater, mainly consisting of Chlorella
(Safafar et al., 2015; Wa
gner et al., 2018). On the contrary, if nu- sorokiniana and Scenedesmus sp. (Supporting Information (SI),
trients are fed in excess, algae can store large quantities of phos- Fig. S1) was used in this study (for further information on the
phorus, e.g., as polyphosphates (Powell et al., 2011), and nitrogen, culture characterization, the reader is referred to Wa
gner et al.
e.g., as protein (Gardner-Dale et al., 2017). Therefore, for the (2016b)). The consortium was cultivated for 21 days on used wa-
particular goal of water resource recovery, the effect of dynamic ter treated by a laboratory scale low-SRT enhanced biological
nitrogen-to-phosphorus molar ratio (NPR) on algal cultivation phosphorus removal system (EBPR) operated as a sequencing batch
needs to be better understood. reactor (SBR) at 3 days SRT. The treated used water was not ster-
A two-stage bacterial-algal cultivation for nutrient removal from ilized prior to feeding the mixed algal consortium. The EBPR was
used water was proposed by Valverde-Pe
rez et al. (2015), whereby fed with used water collected at Mølleåværket WWTP (Kgs. Lyngby,
an enhanced biological phosphorus recovery and removal (EBP2R) Denmark). Details on the operation of the EBPR system can be
process provides optimal cultivation media for green microalgal found elsewhere (Valverde-Pe
rez et al., 2016b).
growth. The combined EBP2R and algal photobioreactor (PBR) is
referred to as TRENS system (Fang et al., 2016). The system pro-
2.1.2. Continuous microalgal cultivation using Chlorella sp.
duces an algal suspension where nutrients are stored in the algal
Two reactors in continuous operation were run for 85 days with
biomass rather than present in the bulk liquid, which can be used
Chlorella sp. (identified based on microscopy, Figs. S1a and SI). The
for fertigation. The TRENS system offers the flexibility of cultivating
culture originated from the mixed consortium described in the
different microalgal species under optimal conditions from a given
previous section. However, Scenedesmus sp. was outcompeted
used water source. In order to keep a stable NPR for algal cultiva-
before the start of this experiment. The culture was fed with treated
tion, Valverde-Pe
rez et al. (2016a) proposed a control structure for
used water collected from a laboratory scale continuous EBPR
EBP2R systems. The study found that, under highly dynamic con-
system operated at 16 days solids retention time, SRT (Figs. S2 and
ditions, the effluent NPR presented some variability around the
SI). More details on the operation of the continuous EBPR can be
optimal ratio. Therefore, understanding the nutrients requirements
rez,
found in the Supporting Information (page 9, SI (Valverde-Pe
and the optimal NPR of microalgal cultivation in dynamic used
2015)). The influent used water fed to the EBPR was taken from
water systems is essential.
Mølleåværket WWTP (Kgs. Lyngby, Denmark).
As used water contains diverse microorganisms, such as algae
and protozoa (Henze et al., 2008), there is high risk of contamina-
tion, especially in open reactors, which may compromise algal 2.2. Microalgal cultivation in 1.4 L continuous PBR
cultivation (Montemezzani et al., 2015). Thus, robust microalgal
species or mixed microalgal consortia are preferred for long term 2.2.1. Continuous reactor operation using the mixed green
reactor operation (Novoveska
et al., 2016). Lynch et al. (2015) sug- microalgal consortium
gest the use of native species, which would outperform other mi- A cylindrical glass PBR with a working volume of 1.4 L was
croorganisms. Moreover, species that can be cultivated in selective operated in the laboratory. The SRT was the same as the hydraulic
environments, such as high salinity or alkalinity, would reduce the residence time (HRT) in the system and it was kept at 2 days.
risk of contamination in such systems. Furthermore, promoting Constant aeration at a flow rate of 30 L/h was used to mix and keep
microalgal growth by optimized cultivation (e.g. sufficient light the microalgae in suspension. CO2 was mixed to the air in a ratio
8e12% to maintain the pH in the 6.5e7.2 range. The reactor was

gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa 3

kept at ambient temperature (23e26  C). Based on previous done using a Motic AE31 microscope (Hong Kong) with a magni-
experience with the mixed culture (Wa
gner et al., 2016b), light in fication of 20x. For image analysis, the software Image pro plus 7 3D
the visible spectrum was continuously supplied from the top of the suite (Media Cybernetics, MD, USA) was used to automate the
reactor at an average intensity of 800 ± 200 mmol photons m2 s1 identification and quantification of the different types of algae
with a custom-built lamp, with a metal-halide bulb (OSRAM©, based on their morphology. The different types of green microalgae
Germany). The reactors were covered with a black cloth around the were distinguished according to their morphology: i.e. Chlorella sp.
wall, to avoid light entering from the side and to mitigate the daily (round and small individual cells, Figs. S1a and SI) and Scenedesmus
light intensity changes that would occur in a laboratory. The treated sp. (elongated cells forming two-to-four-cell colonies, Figs. S1b and
used water, serving as the influent to the PBR, was kept at SI). The diatoms that appeared during cultivation could also be
approximately 5  C. The pH was adjusted to 6.5e7 (with HCl identified based on their morphology (elongated cells, larger than
addition) to avoid phosphorus precipitation in the feed. The NPR of the other two species, Figs. S5a and SI). To distinguish between the
the influent to the PBR was varied during the cultivation. The NPR genera, morphological descriptors (area, diameter, aspect ratio,
in this study is expressed as nitrogen-to-phosphorus molar ratio. which is the ratio between the longest and shortest diameter and
During the first 6.5 days the influent NPR was kept at 17.6 ± 1.6, axis, which describes the long narrow algae) of each type were
selected based on previous experience with the mixed culture automatically acquired using the image analysis software. The
(Wa
gner et al., 2016b). Then, the NPR was lowered to 5.2 until day morphological parameters (Tables S2 and SI) were observed to be
10.5 and finally back to 16.5 ± 0.1 until the end of the operation. The culture condition specific. Therefore, it is suggested to calibrate the
influent phosphorus concentration was kept constant at morphological parameters every time a new consortium is studied.
6 ± 0.7 mg-P/L by adding KH2PO4 (Sigma Aldrich), whilst the The required number of images for the analysis of estimating the
ammonium concentration was adjusted to reach the required NPR distribution of different species was tested on both the mixed
by adding NH4HCO3 (Sigma Aldrich) as nitrogen source. culture (containing Chlorella sorokiniana and Scenedesmus sp.) and
the monoculture (containing Chlorella sp.) by taking up to 50 im-
ages and assessing the cell count and the cell area based on the
2.2.2. Continuous reactor operation using Chlorella sp.
number of images. The required number of images for the mixed
Two 1.4 L glass cylindrical PBRs were used as described in the
culture was found to be 20, whilst for the monoculture it was 10
previous section. The SRT and HRT in the system were kept at 3.5
(Figs. S4 and SI).
days. Constant aeration mixed with CO2 in 30e40% ratio was used
at a flow rate of 20 L/h to adjust the pH in the range of 6.5e7.5 and
mix the suspension. Light was supplied from the top of the reactors 2.4. Batch experiments for model calibration
continuously with a custom-built lamp, providing in average
1262 ± 314 mmol photons m2 s1, with a metal-halide light bulb Three batch experiments were set up using a glass cylinder with
(OSRAM©, Germany). The reactors were covered with a black cloth a working volume of 1.2 L. Continuous lighting was provided by one
around the wall, to avoid light entering from the side of the reactor custom-built lamp with a metal-halide light bulb (OSRAM©, Ger-
(Figs. S3a and SI) and to mitigate the daily light intensity changes many) from the top of the reactor with an average light intensity of
that would occur in a laboratory. The cultivation was done at room 1000 ± 121 mmol photons m2 s1. CO2 enriched airflow was sup-
temperature (approximately 21  C) until a sudden increase in plied from the bottom of the reactor at 20 L/h and at 30e40% CO2 to
ambient temperature due to the summer weather conditions (this guarantee full mixing and to keep pH between 6.5 and 7.5. The
period, from day 45 to day 58 is not discussed in this study). The batch was prepared using 1 L of effluent from the laboratory scale
reactor temperature was then controlled with a cooling system, EBPR system and 200 ml of effluent from the second continuous
whereby cold tap water was circulated around the PBR surface to PBR containing Chlorella sp. The inoculum for the first batch was
lower the temperature inside the reactor to approximately to 21  C taken on day 21, when the culture was adjusted to an NPR of
(Figs. S3b and SI). The treated used water, which served as the 16.8 ± 2. The second batch was inoculated with algae taken on day
influent to the PBR, was kept at approximately 5  C. The pH was 37, when the culture was adjusted to an NPR of 10.3 ± 0.3. Finally,
adjusted to 6.5e7 (with HCl addition) to avoid phosphorus pre- the inoculum for batch 3 was taken on day 79, when the culture
cipitation in the feed. Two identical reactors were operated in was cultivated at 25.8 ± 1.7 NPR. The experiments were run with an
parallel. One reactor was chosen as a reference (Reactor 1 (R1)) and initial NPR of 17. The required nitrogen and phosphorus concen-
was operated at constant influent NPR of 17.3 ± 2. In a second trations were obtained by spiking NaNO3 and KH2PO4 solutions.
reactor (Reactor 2 (R2)) the influent to the PBR was varied. During The initial conditions of the three experiments can be found in
the first 21 days the influent NPR was kept at 16.8 ± 2, then lowered Tables S1 and SI.
to 10.3 ± 0.33 until day 37, and finally back to 17.9 ± 0.5 until day 45.
The NPR in the second reactor was kept at 17.3 ± 3 until day 71, 2.5. Model based assessment, model calibration and statistical
when it was increased to 25.8 ± 1.7 until day 85, the end of the analysis
reactor operation. The nitrogen and ortho-phosphate concentra-
tions were adjusted to reach the required NPR by adding NaNO3 The ASM-A process model, developed in a previous study
(Sigma Aldrich) as nitrogen and KH2PO4 (Sigma Aldrich) as phos-
gner et al., 2016b), was used for the simulations and parameter
(Wa
phorus source. The phosphorus concentration was kept constant at
estimation. ASM-A, implemented in Matlab (The MathWorks,
3.9 ± 0.33 mg-P/L and nitrogen was varied to reach the required
Natick, Massachusetts, USA), can predict the uptake and storage of
NPR.
nitrogen and phosphorus as well as microalgal growth and decay
under photoautotrophic and heterotrophic conditions. Default pa-
2.3. Image analysis for community composition description rameters were used for the mixed culture and model was recali-
brated when assessing the monoculture performance. Parameter
The culture in each experiment was monitored using an inno- estimation was carried out based on the global optimisation
vative image analysis method developed during this study. One method for parameter estimation, Latin Hypercube Sampled priors
drop of suspension was taken with a disposable plastic pipette and for Simplex (LHSS) (W
agner et al., 2016b). 500 simulations were
the droplet was covered with a cover slip. Bright field imaging was found sufficient to reach convergence for parameter estimation.
4
gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa

Values for parameters not estimated in this study were taken from originated from the WWTP and proliferated under the altered
the original ASM-A calibration. The average light intensity in the cultivation conditions. Indeed, diatoms have been reported as
reactor was calculated based on integration of the Lambert-Beer indigenous algae species in sewage streams in Scandinavia
law. However, in this study we used a time variable average light (Krustok et al., 2015). Many diatoms have lower NPR compared to
intensity, by calculating it for each time step and updating it during the mixed culture used in this study (Dang et al., 2018), so lower
the model simulations. The incident light intensity varied during NPR may have selected for their growth. Furthermore, diatoms are
the batch experiments, due to the decrease of the height of the able to grow in waters with very limited nitrogen availability.
suspension in the reactor resulting from the daily sampling. Thus, Indeed, the affinity coefficients reported for these microalgae are
the incident light intensity was updated during the simulations considerably lower compared to those reported for the mixed
(see level of complexity 2 for light modelling in (Wa
gner et al., culture used in this study (ammonia affinity coefficients for di-
2018)). The Janus coefficient was calculated for model evaluation atoms are in the order of 1-103 mg-N/L, Fan et al., 2003, compared
(Sin et al., 2008). The model parameter estimates are validated to 4-10 mg-N/L for our mixed culture, W
agner et al., 2016b), which
when the Janus coefficient approaches 1. make them efficient k-strategists. Alternatively, diatoms build
synergies with cyanobacteria in low nutrient environments,
whereby cyanobacteria fix atmospheric nitrogen and make it
2.6. Analytical methods and calculations
bioavailable for diatoms (Foster et al., 2011). Considering these
aspects, when decreasing the ammonium in the influent to
Biomass in the continuous and batch reactors was analysed by
decrease the NPR, diatoms were still able to grow in the PBR, while
measuring the total suspended solids (TSS) using glass fibre filter
the mixed culture’s growth was limited. When NPR was above 16,
(Advantec©, USA) with a pore size of 0.6 mm (APHA et al., 1999).
nitrogen was not a limiting growth factor, thus promoting fast
Total COD, total nitrogen and total phosphorus measurements in
growing green microalgae (e.g., Chlorella sp., which are r-strate-
the suspension were done using commercial test kits (Hach-
s et al., 2019). The amount of diatoms increased to up to
gists; Gale
Lange©, CO, USA). Following sample filtration (0.2 mm filter),
8% of total cell count by day 10. Their relative abundance was much
ammonium, nitrate, nitrite and phosphate concentrations were
larger when accounting for the cell area (up to 34%), due to their
measured using test kits supplied by Merck© (NJ, USA) and soluble
3e5 times higher cell size relative to Chlorella sp. and Scenedesmus
COD was measured using Hach-Lange© test kits (CO, USA). The COD
sp. The cell area is relevant as it can be related to the TSS of the
of the microalgal biomass was calculated by the difference of the
biomass (Figs. S6 and SI), and thus it can allow the approximation of
total and soluble COD. The TSS to COD conversion factor was esti-
the mass fraction of the different species. Thus, although the cell
mated to be in average 0.9 (calculated as described in SI). The in-
count of diatoms was low compared to the other species, diatoms
ternal cell quota of nitrogen was calculated based on the difference
constituted a significant fraction of the biomass concentration.
of total nitrogen in the algal suspension (algae þ medium) and total
Moreover, the number and the size of ciliates increased in the
soluble nitrogen in the filtrate (soluble organic N, NHþ 
4 , NO2 and
reactor (Figs. S5b and SI), yielding a significant increase of the
NO 3 ). The internal cell quota of phosphorus was obtained by the
relative cell area of other non-classified species (66%, Fig. 1).
difference of total phosphorus in the algal suspension and soluble
However, ciliates were quickly washed out at day 10 (Fig. 1). Di-
phosphate in the filtrate. Incoming light intensity and pH were
atoms contain several metabolites that can have a detrimental ef-
monitored using LI-1400 Data Logger with LI-193 Spherical Un-
fect on the reproduction of ciliates (Miralto et al., 1999). Thus, the
derwater Quantum Sensor (LI-COR, USA) and Multi 3430 Digital pH
proliferation of diatoms in the PBR served to control ciliates pop-
meter for pH-Electrode Sentix 940 sensor (WTW, Germany),
ulation, which fed on them, and eventually led to lower numbers of
respectively. The pigment content of the biomass was measured as

gner et al. (2018). Chlorophyll a and b as well as these predators once the NPR was restored. The presence of ciliates
described in Wa
and other grazers act as selective factor promoting the co-existence
some carotenoids (lutein, b-carotene, violaxanthin) were targeted
of Chlorella sp. and Scenedesmus sp., as the second is more resilient
during the analysis.
to predators despite its comparably lower growth rates (Gale s et al.,
Protein content was calculated by multiplying the nitrogen
2019). On day 8, the fraction of Chlorella sp. started to increase
quota by a nitrogen-to-protein factor as suggested by Gardner-Dale
while Scenedesmus sp. decreased. Nevertheless, by day 10, their cell
et al. (2017). Nitrogen-to-protein conversion factor value was taken
count was reduced, possibly due to the presence of grazers that
from Templeton and Laurens (2015). For Chlorella sp. it is reported
prefer Chlorella over Scenedesmus sp. due to their smaller size
to be 5.04 g protein/g N.
(Borowitzka and Moheimani, n.d.). After the NPR was set back to
16.5 ± 0.1, the diatoms were outcompeted from the system, sug-
3. Results and discussion gesting that the cultivation conditions were more optimal for
Chlorella and Scenedesmus sp. Interestingly, the relative ratio of
3.1. Impact of varying nutrient availability on microalgal diversity Chlorella and Scenedesmus sp. had shifted from day 11e21, reaching

ntara
77% of Chlorella sp. of the total cell count at day 21 (Fig. 1). Alca
3.1.1. Mixed culture dynamics et al. (2015) similarly observed that Scenedesmus sp. were initially
The image analysis tool was used to monitor the culture more abundant, whilst by the end of the cultivation period Chlorella
composition in the mixed microalgal consortium. At the beginning sp. proliferated. Thus, it is hypothesised that the cultivation con-
of the cultivation the mixed microalgal consortium contained ditions are more optimal for Chlorella sp. than Scenedesmus sp. in
mostly Scenedesmus sp., about 83% of the total cell count (Fig. 1). our case, also supported by Beuckels et al. (2015), who found that
Chlorella sp. were present at 9% along with some other species Chlorella sp. is capable of accumulating more nitrogen than Scene-
belonging mainly to phylum Chlorophyta and some smaller ciliates desmus sp. This contradicts with other studies which showed that
at 8%. The composition did not change in the first period of the Scenedesmus sp. prefer larger NPR than Chlorella sp. when phos-
cultivation, when the NPR was 17.6 ± 1.6. When the influent NPR phorus concentration is varied (Marcilhac et al., 2015).
was lowered to 5.2, there was a sudden appearance of diatoms
identified as Nitzschia sp. based on microscopic observation
(Figs. S5a and SI). These diatoms were seeded from the influent 3.1.2. Monoculture dynamics
water to the PBR (based on microscopic observations) that probably A monoculture of Chlorella sp. was cultivated in continuous

gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa 5

Fig. 1. Variation in the culture composition during the 21 days cultivation period as affected by the NPR (indicated on the top of the graphs). (a) The cell counts are presented as the
fraction of the total cell count. (b) The cell area is presented as the fraction of the total cell area (unit in pixels). The total cell count (number of cells) and the total cell area (total
number of pixels per field of view) is presented on the figures, suggesting how the culture density changed.

Fig. 2. Total cell counts (number of cells) (a) and cell area (unit in pixels) (b) of the Chlorella sp. during continuous reactor operation as affected by the NPR (indicated on the top of
the graphs). The gap in the x-axis refers to the period where temperature control was not applied and thus it is not considered. Reactor 1 refers to the reference reactor operated at
17.3 ± 2 NPR whilst Reactor 2 refers to the reactor with varying NPRs.

reactor operation using effluent water from the upstream EBPR operational period. We note that only Chlorella sp. were detected
process. The culture composition did not change throughout the 85 during this operational period. Grazers were not present in the
days of cultivation and Chlorella sp. remained as a single microalgal influent and therefore there was not competitive advantage for
species even though we used treated used water from the EBPR Scenedesmus sp., which were outcompeted by Chlorella sp. due to
system without disinfection. There was some variation in the cell their comparably higher growth rates (Gale s et al., 2019). The main
number and cell area during the cultivation (Fig. 2) that could be difference between the two operational periods is the influent.
related to the variation in biomass concentration. However, the When growing the mixed culture, the effluents were collected from
correlation with the TSS concentration is scattered and does not a lab-scale short SRT EBPR system suffering extreme filamentous
show strong relation (Figs. S7 and SI). No major differences were bulking (Valverde-Pe
rez et al., 2016b), while in the second case
observed in the cell counts from both PBR, suggesting the micro- effluents were collected from a lab-scale EBPR system exhibiting
algae grew at similar rates in both of them. This is likely a conse- good performance (Valverde-Pe
rez, 2015). Poor settleability pro-
quence of excessive volumetric nutrient loading, which allowed motes the proliferation of swimming protozoa (Liu et al., 2008),
Chlorella sp. to grow under non-limiting conditions. Indeed, only at which explains why predators were only present in the study when
the end of the operational period, when the NPR was set to the mixed culture was grown. In this experiment, the lowest NPR
10.3 ± 0.3, nitrate reached levels close to 0 mg-N/L, suggesting that was higher than in case of the mixed culture experiment. NPR for
nitrate was not limiting during most of the operational period and Chlorella sp. biomass composition (Beuckels et al., 2015) are typi-
Chlorella sp. grew at maximum capacity during most of the cally higher than for diatoms (Garcia et al., 2018) and therefore, a
6
gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa

proper control of the NPR could effectively keep Chlorella sp. cul- g-TSS and was kept stable during the operational period, suggesting
ture stable in the PBR. Thus controlling the NPR is a powerful tool to nitrogen was the limiting factor for biomass growth. Similar to our
regulate and stabilize PBR used for resource recovery in combina- previous observations with the mixed culture, the nitrogen quota
tion with bacterial systems designed for carbon capture at low SRTs, shows high variability (Wa
gner et al., 2016b). Interestingly, the
as it is the case for the TRENS system (Valverde-Pe
rez et al., 2016a). pigments concentrations were not considerably affected by the
These results also demonstrate image analysis as a powerful tool change of NPR (Figs. S9 and SI). Nevertheless, chlorophyll concen-
to monitor algal system performance. This is especially relevant for tration showed high variability until day 10, which made it difficult
open pond systems, which are usually subjected to contamination to demonstrate that reduced NPR yielded low chlorophyll content.
by algae and other microbes (Montemezzani et al., 2015). Similar It should be noted that once the NPR was restored in day 10.5,
tools have been designed for activated sludge systems to, e.g., chlorophyll content was more stable and always above 2.5 mg-
monitor filamentous bulking, and successfully implemented in full- chlorophyll/g-TSS. Previous studies demonstrated that under low
scale systems (Mesquita et al., 2013). However, the application of nitrogen feeding conditions, such as those when the PBR was
these tools for microalgal cultures is limited to cell counting and operated with NPR ¼ 5.2, can yield to lower chlorophyll content
morphology characterization, without distinguishing among spe- (Wa
gner et al., 2018).
cies in mixed cultures (Havlik et al., 2013). More complex moni- In the cultivation of the Chlorella sp. monoculture, no significant
toring tools working at different wavelengths can perform similar difference was observed in the TSS concentration (Fig. 4a) in R1 and
measurements to those presented in our work (Winckelmann et al., R2, except for the NPR ¼ 10.3 ± 0.3 for R2, where biomass con-
2016). However, to the best of our knowledge, no studies have used centration was slightly lower in R2 (0.09 g/L) by the end of the
field microscopy combined with image analysis tools to charac- period. Compared to the reference reactor with constant nutrient
terize microalgal composition in mixed cultures. The method supply (R1), in R2, during the period of decreased nutrient avail-
developed in this study has the potential to be implemented in full- ability (NPR ¼ 10.3 ± 0.3), the nitrate concentration decreased until
scale algal systems, offering new controlled variables, e.g. algal a steady-state at 1 mg N/L (Fig. 4d), lower than R1, whilst the in-
diversity, thereby leading to the development of innovative control ternal nitrogen cell quota decreased to 0.01 gN/gCOD below that of
strategies beyond conventional pH or cell density control (Olivieri R1 in the beginning of the low NPR period (Fig. 4b). During the
et al., 2014). phase of excess nutrient supply (NPR ¼ 25.8 ± 1.7), there was an
increase of the nitrate concentration in R2 until a steady state at
23 mg N/L, higher than R1, and an increase of the internal cell quota
3.2. Impact of influent NPR on treatment efficiency and nutrient (0.12 gN/gCOD), higher than R1. This is reflected in the removal of
recovery nitrate (Fig. 4f). On average there was 75% nitrogen removal that
increased to 95% under 10.3 ±.3 NPR and decreased to 60% under
The nutrients removal and effluent quality of the PBRs were 25.8 ± 1.7 NPR. The phosphate and the internal phosphorus quota
assessed during the cultivation of mixed microalgal consortium (Fig. 4 e and c) showed no difference between the two reactor
(Fig. 3, Figs. S8 and SI) and through the cultivation of the Chlorella operations, resulting in a 60% average removal (Fig. 4f). Since
sp. monoculture. As for the cultivation of the mixed microalgal phosphorus was not fully removed during the experiment, prob-
consortium, the phosphate removal decreased significantly, from ably, the phosphorus load was too high and SRT and HRT too low
95% to 40%, as a result of decreasing influent NPR, likely due to during this experiment, thus limiting P removal. Nitrate accumu-
nitrogen limitation. Once the cultivation conditions were restored, lated in the last period, suggesting that feeding load was above the
the removal for both nitrogen and phosphorus reached up to 95%. optimal range of NPR. The pigment content was not significantly
The effluent quality was significantly affected by the changes in affected by the change of NPR as no difference was observed be-
microalgal composition as a result of changes in influent NPR tween R1 and R2 (Figs. S10 and SI), possibly because algae were not
(Figs. S8 and SI). The biomass concentration and the soluble effluent nutrient starved long enough.
nitrogen concentration decreased during the NPR ¼ 5.2 period, In the beginning of the operating period with influent
while the effluent soluble phosphate concentration increased NPR ¼ 10.3 ± 0.3, a decrease of nitrogen quota in R2 was observed,
(Figs. S8 and SI). Phosphorus quota reached a maximum of 0.15 g-P/ whilst TSS concentration remained unchanged. This result suggests
that algae were able to grow under nitrogen limiting conditions at
the expense of their internal nitrogen quota. In the second half of
the NPR ¼ 10.3 ± 0.3 period, the biomass concentration decreased
in R2 while the nitrogen quota replenished. Hence, we conclude
that the growth rate may not be affected by changes in the influent
NPR lasting for only a few days. Thus, the control by Valverde-Pe
rez
et al. (2016a) will allow stable and optimal algal cultivation.
Furthermore, sub-optimal NPR ratios may not result in growth rate
limitation as long as both nitrogen and phosphorus are fed in-
excess concentrations at a given SRT, as suggested by Arbib et al.
(2013). However, if one of the nutrients becomes limiting for the
given operational conditions, algal biomass productivity is
considerably reduced. The NPR inside the algae was calculated
(Fig. 5a) and it was found to be varying between 2.7 and 21.4 mol N/
mol P. This range of NPR was obtained when nitrogen was varied
and might expand when phosphorus is varied as well. The calcu-
lated protein content of the algae (calculated by direct conversion
from the nitrogen quota) is between 0.06 and 0.8 g protein/g DW
Fig. 3. Removal of ammonium and phosphorus in the mixed microalgal consortium
during the 21 days of cultivation (expressed as fraction of influent nutrients present in (Fig. 5b) and it is around 0.4 g protein/g DW under optimal NPR
the effluent). The NPR present in the system is indicated above the figure. The NPR operation. This is in agreement with literature where Rasouli et al.
(17.6 ± 1.6, 5.2 and 16.5 ± 0.1) is shown for each period.

gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa 7

Fig. 4. Biomass concentration (a), internal nitrogen quota (b), internal phosphorus quota (c), bulk nitrate concentration (d), bulk phosphate concentration (e), and removal of
nitrogen and phosphorus (f) during the cultivation of Chlorella sp. in used water resources in Reactor 1 and 2. Reactor 1 refers to the reference reactor operated at 17.3 ± 2 NPR whilst
Reactor 2 refers to the reactor with varying NPRs.

Fig. 5. Nitrogen e to e phosphorus ratio (NPR) calculated inside the algae (a) and protein content of the algae calculated (b) during the cultivation of Chlorella sp. in used water
resources in Reactor 1 and 2.

(2018) reported 45% protein content for Chlorella sorokiniana and 3.3. Effect of influent NPR on microalgal growth kinetics
Molazadeh et al. (2019) reported between 43 and 61% protein
content for Chlorella vulgaris. The high protein content is, indeed, 3.3.1. Changes in microalgal culture composition
comparable to other microbial protein sources, such as methano- To assess the impact of alteration in the influent nutrient supply
trophic bacteria (Valverde-Pe
rez et al., 2020). to PBRs, the average parameter set reported in Wa
gner et al.
(2016b) was used to simulate the results obtained in the
8
gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa

Fig. 6. Simulation results of the mixed microalgal species cultivation. The red vertical dashed lines represent the time when the NPR was changed. Simulation 1 (blue line)
represents the simulation of the whole cultivation period. Discrepancies following the decrease of NPR are due to the change in culture composition. Simulation 2 (orange line)
represents the simulation of the second 16.5 ± 0.1 NPR period. Discrepancies between blue and red lines are due to initial conditions after the second NPR shift, which are given by
the simulation model for the full cultivation period simulation and set to the experimental values in simulation 2. (For interpretation of the references to colour in this figure legend,
the reader is referred to the Web version of this article.)

continuous reactor operation with the mixed consortium. When Table 1

only Chlorella and Scenedesmus sp. were present in the culture, Estimated parameter values obtained for the three batch experiments. The values
are presented as mean ± standard deviation. The 17-17 batch denotes the cultivation
accurate prediction of the biomass concentration, the nitrogen and
where the inoculum was taken after the 16.8 ± 2 NPR cultivation. The 10e17 batch
phosphorus internal cell quota and the soluble nitrogen and denotes the cultivation where the inoculum was taken after the 10.3 ± 0.33 NPR
phosphorus species were obtained (Fig. 6). When the influent NPR cultivation. The 25-17 batch denotes the cultivation where the inoculum was taken
was lowered to 5.2 and diatoms and other microbes proliferated in after the 25.8 ± 1.7 NPR cultivation.
the culture, the simulations failed to predict the measured data. 17-17 batch 10-17 batch 25-17 batch
This discrepancy between measured and predicted results indicates
mA,max (d1) 2.66 ± 0.1 3.4 ± 0.8 2.27 ± 0.4
altered process kinetics as a result of the diatom invasion, and the KNO,Alg (gN m3) 14.55 ± 1 13 ± 1.4 13.7 ± 1.8
consequent deterioration of nutrient recovery in the PBR. Some KPO,Alg (gP m3) 4.48 ± 0.7 4 ± 0.58 3.7 ± 0.5
discrepancies exist for the internal nitrogen quota, which can be kNO,Alg (gN g1COD d1) 0.14 ± 0.02 0.1 ± 0.02 0.14 ± 0.05
consequence of a potential change in the nitrogen storage kinetics kPO,Alg (gP g1COD d1) 0.043 ± 0.007 0.023 ± 0.003 0.045 ± 0.005
bAlg (d1) 0.45 ± 0.04 0.31 ± 0.12 0.38 ± 0.1
after the nitrogen limitation. The latter phenomenon was also
observed by Wa
gner et al. (2016b).

Using the estimated parameter set obtained in the batch that

3.3.2. Model calibration and evaluation of the operation with
was conducted after the first period with NPR ¼ 16.8 ± 2 (17-17
Chlorella sp.
batch, Table 1) we assessed the model prediction accuracy of the
The batch reactors were used to estimate kinetic parameters of
other two batches using the Janus coefficient (Tables S3 and SI). We
the ASM-A model (Fig. S11 - S13, SI). The half-saturation coefficient
found J~1, suggesting that the parameter set was fit to predict the
and maximum uptake rate of nitrogen and phosphorus (KNO,Alg,
PBR performance in the full operation period. Indeed, comparably
KPO,Alg, kNO,Alg and kPO,Alg),the maximum specific growth rate (mA,max)
good fit is obtained with the simulation results throughout the 85
and the biomass decay rate (bAlg) were estimated using the LHSS
days of experiments (Fig. 7) in terms of microalgal biomass con-
method. Since the ammonium concentration in the PBR was low
centration, soluble nitrate and phosphate concentration and in-
(below 0.1 mg/L) during the reactor operation as nitrate was the
ternal phosphorus quota. The decrease of internal nitrogen quota in
nitrogen source, KNH,Alg and kNH,Alg were not estimated in this
the beginning of the NPR ¼ 10.3 ± 0.3 period was not captured by
experiment and literature values from Wa
gner et al. (2016b) were
the model prediction. In the same period, biomass concentration is
used. The estimated parameter subsets do not show significant
slightly over-predicted, possibly due to error propagation from the
differences obtained for the three batch experiments (Table 1).

gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa 9

Fig. 7. Simulation of Reactor 2 with varying NPR. The blue line indicates the model prediction of the measured data. (For interpretation of the references to colour in this figure
legend, the reader is referred to the Web version of this article.)

internal nitrogen quota predictions. This was the only period where nitrogen concentration and the internal nitrogen quota. The ASM-A
algae suffered nitrogen limitation. When dynamics expose algae to model was calibrated for this monoculture and model simulations
nitrogen limitation conditions the maximum nitrate uptake rate could predict the measurement data in continuous operation using
(kNO,Alg) shows variability (Wa
gner et al., 2016b) and therefore a a single parameter subset. Under nutrient limitation conditions,
recalibration of that parameter would have been needed for proper which affects the nitrate uptake rates, the model could not predict
prediction of that period. When the algae was not under stress the nitrogen quota.
condition, no change in the kinetics for the monoculture is This study demonstrates that NPR is a powerful control
observed. Taken together, the ASM-A simulation model can predict parameter of the PBR performance. Moreover, process configura-
PBR performance under dynamic conditions as long as there is no tions or control schemes reported in literature (see, e.g. Valverde-
extreme nutrient limitation, which heavily affects nitrate uptake Pe
rez et al. (2016a)) could effectively optimize the PBR
rates or no changes the microbial composition of the consortium performance.
which affects the biokinetics.
Credit author statement
4. Conclusions
Dorottya S. W
agner, Conceptualization, Methodology, Investi-
Effects of varying NPR on microalgal cultivation in terms of gation, Formal analysis, Software, Data curation, Visualization,
microbial composition and process performance were assessed Writing - original draft, Clarissa Cazzaniga, Conceptualization,
when cultivated on used water resources. During the cultivation of Investigation, Data curation. Michael Steidl, Conceptualization,
the mixed microalgal species, diatoms, an indigenous algae, Investigation, Data curation. Arnaud Dechesne, Methodology, Su-
proliferated in the reactor when the NPR was lowered below the pervision, Formal analysis, Writing - review & editing. Borja
Valverde-Pe
rez, Conceptualization, Investigation, Software, Writing
optimal NPR range, and they were outcompeted once the NPR was
restored. Changes in microbial community could be effectively - review & editing. Benedek Gy. Plo
sz, Conceptualization, Super-
tracked by an image analysis method, which has the potential of vision, Writing - review & editing.
being a monitoring tool for full-scale algal systems. The phosphate
removal decreased significantly, after the NPR was lowered in the Declaration of competing interest
influent but once the cultivation conditions were restored for the
seeded green microalgal species, the removal for both nitrogen and The authors declare that they have no known competing
phosphorus recovered. Biomass and nitrogen concentration financial interests or personal relationships that could have
decreased while phosphorus concentration increased in the appeared to influence the work reported in this paper.
effluent by decreasing the NPR. The ASM-A model could capture the
measured data under optimal NPR operation, but it failed to predict Acknowledgements
the measured results when diatoms proliferated under low NPR.
A monoculture of Chlorella sp. cultivated on used water re- Dorottya W
agner thanks the European Commission, (E4WATER
sources with varying NPR remained stable and contamination-free. Project, FP7-NMP-2011.3.4-1 grant agreement 280756) for the
Low NPR condition decreased the biomass concentration, effluent funding. Borja Valverde-Pe
rez thanks the Integrated Water
10
gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa

Technology (InWaTech) project (http://www.inwatech.org) for the microalgae and its biochemical basis. Eur. J. Phycol. 37 (1), 1e17. https://doi.org/
10.1017/S0967026201003456.
financial support.
Geiss, R.H., Rice, K.P., Keller, R.R., Chen, G., Zhao, L., Qi, Y., 2015. Enhancing the
productivity of microalgae cultivated in wastewater toward biofuel production:
Appendix A. Supplementary data a critical review. Appl. Energy 137, 282e291.
Havlik, I., Lindner, P., Scheper, T., Reardon, K.F., 2013. On-line monitoring of large
cultivations of microalgae and cyanobacteria. Trends Biotechnol. 31, 406e414.
Supplementary data to this article can be found online at https://doi.org/10.1016/j.tibtech.2013.04.005.
https://doi.org/10.1016/j.chemosphere.2020.127939. Henze, M., Van Loosdrecht, M.C.M., Ekama, G.A., Brdjanovic, D., 2008. Biological
Wastewater Treatment: Principles, Modelling and Design. IWA Publishing, UK.
Ikaran, Z., Sua
rez-Alvarez, S., Urreta, I., Castan
~o
n, S., 2015. The effect of nitrogen
References limitation on the physiology and metabolism of Chlorella vulgaris var L3. Algal
Res 10, 134e144.

ntara, C., Domínguez, J.M., García, D., Blanco, S., Pe
Alca
rez, R., García-Encina, P.A., Krustok, I., Odlare, M., Shabiimam, M.A., Truu, J., Truu, M., Ligi, T., Nehrenheim, E.,
Mun ~ oz, R., 2015. Evaluation of wastewater treatment in a novel anoxic-aerobic 2015. Characterization of algal and microbial community growth in a waste-
algal-bacterial photobioreactor with biomass recycling through carbon and water treating batch photo-bioreactor inoculated with lake water. Algal Res 11,
nitrogen mass balances. Bioresour. Technol. 191, 173e186. 421e427. https://doi.org/10.1016/j.algal.2015.02.005.
Anbalagan, A., Schwede, S., Lindberg, C.F., Nehrenheim, E., 2016. Influence of hy- Liu, J., Vyverman, W., 2015. Differences in nutrient uptake capacity of the benthic
draulic retention time on indigenous microalgae and activated sludge process. filamentous algae Cladophora sp., Klebsormidium sp. and Pseudanabaena sp.
Water Res. 91, 277e284. under varying N/P conditions. Bioresour. Technol. 179, 234e242.
APHA, American Water Works Association, Water Environment Federation, 1999. Liu, J., Yang, M., Qi, R., An, W., Zhou, J., 2008. Comparative study of protozoan
Standard Methods for the Examination of Water and Wastewater. Standard communities in full-scale MWTPs in Beijing related to treatment processes.
Methods, Washington DC. Water Res. 42 https://doi.org/10.1016/j.watres.2007.11.020, 1907e1918.
Arbib, Z., Ruiz, J., Alvarez-Diaz, P., Garrido-Perez, C., Barragan, J., Perales, J.A., 2013. Lynch, F., Santana-Sa
nchez, A., Ja
€msa €, M., Sivonen, K., Aro, E.M., Allahverdiyeva, Y.,
Photobiotreatment: influence of nitrogen and phosphorus ratio in wastewater 2015. Screening native isolates of cyanobacteria and a green alga for integrated
on growth kinetics of Scenedesmus obliquus. Int. J. Phytoremediation 15, wastewater treatment, biomass accumulation and neutral lipid production.
774e788. Algal Res 11, 411e420.
Batstone, D.J., Hülsen, T., Mehta, C.M., Keller, J., 2015. Platforms for energy and Marcilhac, C., Sialve, B., Pourcher, A.M., Ziebal, C., Bernet, N., Be
line, F., 2015. Control
nutrient recovery from domestic wastewater: a review. Chemosphere 140, of nitrogen behaviour by phosphate concentration during microalgal-bacterial
2e11. https://doi.org/10.1016/j.chemosphere.2014.10.021. cultivation using digestate. Bioresour. Technol. 175, 224e230.

chet, Q., Shilton, A., Guieysse, B., 2016. Maximizing productivity and reducing
Be Mata, T.M., Martins, A.A., Caetano, N.S., 2010. Microalgae for biodiesel production
environmental impacts of full-scale Algal production through optimization of and other applications: a review. Renew. Sustain. Energy Rev. 14, 217e232.
open pond depth and hydraulic retention time. Environ. Sci. Technol. 50, Matassa, S., Batstone, D.J., Hülsen, T., Schnoor, J., Verstraete, W., 2015. Can direct
4102e4110. conversion of used nitrogen to new feed and protein help feed the world?
Beuckels, A., Smolders, E., Muylaert, K., 2015. Nitrogen availability influences Environ. Sci. Technol. 49, 5247e5254.
phosphorus removal in microalgae-based wastewater treatment. Water Res. 77, Mayers, J.J., Flynn, K.J., Shields, R.J., 2014. Influence of the N: P supply ratio on
98e106. biomass productivity and time-resolved changes in elemental and bulk
Boelee, N.C., Temmink, H., Janssen, M., Buisman, C.J.N., Wijffels, R.H., 2011. Nitrogen biochemical composition of Nannochloropsis sp. Bioresour. Technol. 169,
and phosphorus removal from municipal wastewater effluent using microalgal 588e595.
biofilms. Water Res. 45, 5925e5933. Mehta, C.M., Khunjar, W.O., Nguyen, V., Tait, S., Batstone, D.J., 2015. Technologies to
Borowitzka, M.A., Moheimani, N.R., 2013, n.d. Algae for Biofuels and Energy. recover nutrients from waste streams: a critical review. Crit. Rev. Environ. Sci.
Cai, T., Park, S.Y., Li, Y., 2013. Nutrient recovery from wastewater streams by Technol. 45, 385e427.
microalgae: status and prospects. Renew. Sustain. Energy Rev. 19, 360e369. Mesquita, D.P., Amaral, A.L., Ferreira, E.C., 2013. Activated sludge characterization
https://doi.org/10.1016/j.rser.2012.11.030. through microscopy: a review on quantitative image analysis and chemometric
Coppens, J., Grunert, O., Van Den Hende, S., Vanhoutte, I., Boon, N., Haesaert, G., De techniques. Anal. Chim. Acta 802, 14e28. https://doi.org/10.1016/
Gelder, L., 2016. The use of microalgae as a high-value organic slow-release J.ACA.2013.09.016.
fertilizer results in tomatoes with increased carotenoid and sugar levels. Miralto, A., Barone, G., Romano, G., Poulet, S.A., Ianora, A., Russo, G.L., Buttino, I.,
J. Appl. Phycol. 28, 2367e2377. https://doi.org/10.1007/s10811-015-0775-2. Mazzarella, G., Laablr, M., Cabrini, M., Glacobbe, M.G., 1999. The insidious effect
Dang, H., Hall, E., Lovejoy, C., Ptacnik, R., Yvon-Durocher, G., Martiny, A.C., of diatoms on copepod reproduction. Nature 402, 173e176. https://doi.org/
Garcia, N.S., Sexton, J., Riggins, T., Brown, J., Lomas, M.W., 2018. High variability 10.1038/46023.
in cellular stoichiometry of carbon, nitrogen, and phosphorus within classes of Molazadeh, M., Danesh, S., Ahmadzadeh, H., Pourianfar, H.R., 2019. Influence of CO2
marine eukaryotic phytoplankton under sufficient nutrient conditions. Front. concentration and N:P ratio on Chlorella vulgaris-assisted nutrient bioreme-
Microbiol. 9, 543. https://doi.org/10.3389/fmicb.2018.00543. diation, CO2 biofixation and biomass production in a lagoon treatment plant.
De Francisci, D., Su, Y., Iital, A., Angelidaki, I., 2018. Evaluation of microalgae pro- J. Taiwan Inst. Chem. Eng. 96, 114e120.
duction coupled with wastewater treatment. Environ. Technol. 39, 581e592. Montemezzani, V., Duggan, I.C., Hogg, I.D., Craggs, R.J., 2015. A review of potential
Dickinson, K.E., Whitney, C.G., McGinn, P.J., 2013. Nutrient remediation rates in methods for zooplankton control in wastewater treatment High Rate Algal
municipal wastewater and their effect on biochemical composition of the Ponds and algal production raceways. Algal Res 11, 211e226.
microalga Scenedesmus sp. AMDD. Algal Res. 2, 127e134. Novoveska
, L., Franks, D.T., Wulfers, T.A., Henley, W.J., 2016. Stabilizing continuous
Fan, C., Glibert, P.M., Burkholder, J.A.M., 2003. Characterization of the affinity for mixed cultures of microalgae. Algal Res 13, 126e133.
nitrogen, uptake kinetics, and environmental relationships for Prorocentrum Olivieri, G., Salatino, P., Marzocchella, A., 2014. Advances in photobioreactors for
minimum in natural blooms and laboratory cultures. Harmful Algae 2, intensive microalgal production: configurations, operating strategies and ap-
283e299. https://doi.org/10.1016/S1568-9883(03)00047-7. plications. J. Chem. Technol. Biotechnol. 89, 178e195. https://doi.org/10.1002/
Fang, L.L., Valverde-Pe
rez, B., Damgaard, A., Plo
sz, B.G., Rygaard, M., 2016. Life cycle jctb.4218.
assessment as development and decision support tool for wastewater resource Powell, N., Shilton, A., Pratt, S., Chisti, Y., 2011. Luxury uptake of phosphorus by
recovery technology. Water Res. 88, 538e549. microalgae in full-scale waste stabilisation ponds. Water Sci. Technol. 63,
Foster, R.A., Kuypers, M.M.M., Vagner, T., Paerl, R.W., Musat, N., Zehr, J.P., 2011. Ni- 704e709.
trogen fixation and transfer in open ocean diatom-cyanobacterial symbioses. Rasouli, Z., Valverde-Pe
rez, B., D’Este, M., De Francisci, D., Angelidaki, I., 2018.
ISME J. 5, 1484e1493. https://doi.org/10.1038/ismej.2011.26. Nutrient recovery from industrial wastewater as single cell protein by a co-
s, A., Bonnafous, A., Carre
Gale
, C., Jauzein, V., Lanougue re, E., Le Floc’h, E., Pinoit, J., culture of green microalgae and methanotrophs. Biochem. Eng. J. 134,
Poullain, C., Roques, C., Sialve, B., Simier, M., Steyer, J.P., Fouilland, E., 2019. 129e135. https://doi.org/10.1016/J.BEJ.2018.03.010.
Importance of ecological interactions during wastewater treatment using High Redfield, A.C., 1958. The biological control of chemical factors in the environment.
Rate Algal Ponds under different temperate climates. Algal Res 40, 101508. Am. Sci. 46, 205e221.
https://doi.org/10.1016/j.algal.2019.101508. Rhee, G.-Y., 1978. Effects of N:P atomic ratios nitrate limitation on algal growth, cell
Gao, H., Scherson, Y.D., Wells, G.F., 2014. Towards energy neutral wastewater composition, nitrate uptake. Limnol. Oceanogr. 23, 10e25.
treatment: methodology and state of the art. Environ. Sci. Process. impacts 16, Rhee, G.-Y., Gotham, I.J., 1980. Optimum N:P ratios and coexistence of planktonic
1223e1246. algae. J. Phycol. 16, 486e489.
Garcia, N.S., Sexton, J., Riggins, T., Brown, J., Lomas, M.W., Martiny, A.C., 2018. High Safafar, H., Wagenen, J. Van, Møller, P., Jacobsen, C., 2015. Carotenoids, phenolic
variability in cellular stoichiometry of carbon, nitrogen, and phosphorus within compounds and tocopherols contribute to the antioxidative properties of some
classes of marine eukaryotic phytoplankton under sufficient nutrient condi- microalgae species grown on industrial wastewater. Mar. Drugs 13, 7339e7356.
tions. Front. Microbiol. 9, 1e10. Sforza, E., Ramos-Tercero, E.A., Gris, B., Bettin, F., Milani, A., Bertucco, A., 2014.
Gardner-Dale, D.A., Bradley, I.M., Guest, J.S., 2017. Influence of solids residence time Integration of Chlorella protothecoides production in wastewater treatment
and carbon storage on nitrogen and phosphorus recovery by microalgae across plant: from lab measurements to process design. Algal Res 6, 223e233.
diel cycles. Water Res. 121, 231e239. https://doi.org/10.1016/ Sin, G., De Pauw, D.J.W., Weijers, S., Vanrolleghem, P.A., 2008. An efficient approach
J.WATRES.2017.05.033. to automate the manual trial and error calibration of activated sludge models.
Geider, R.J., La Roche, J., 2002. Redfield revisited: variability of C:N:P in marine Biotechnol. Bioeng. 100, 516e528. https://doi.org/10.1002/BIT.21769.

gner et al. / Chemosphere 262 (2021) 127939
D.S. Wa 11

Solovchenko, A., Verschoor, A.M., Jablonowski, N.D., Nedbal, L., 2016. Phosphorus Van Den Hende, S., Carre
, E., Cocaud, E., Beelen, V., Boon, N., Vervaeren, H., 2014.
from wastewater to crops: an alternative path involving microalgae. Biotechnol. Treatment of industrial wastewaters by microalgal bacterial flocs in sequencing
Adv. 34 (5), 550e564. https://doi.org/10.1016/j.biotechadv.2016.01.002. batch reactors. Bioresour. Technol. 161, 245e254.
Sutherland, D.L., Turnbull, M.H., Craggs, R.J., 2014. Increased pond depth improves Verstraete, W., Vlaeminck, S.E., 2011. ZeroWasteWater: short-cycling of wastewater
algal productivity and nutrient removal in wastewater treatment high rate algal resources for sustainable cities of the future. Int. J. Sustain. Dev. World Ecol. 18,
ponds. Water Res. 53, 271e281. 253e264.
Templeton, D.W., Laurens, L.M.L., 2015. Nitrogen-to-protein conversion factors Wa
gner, D.S., Radovici, M., Smets, B.F., Angelidaki, I., Valverde-Pe
rez, B., Plo

sz, B.G.,
revisited for applications of microalgal biomass conversion to food, feed and 2016a. Harvesting microalgae using activated sludge can decrease polymer
fuel. Algal Res 11, 359e367. https://doi.org/10.1016/J.ALGAL.2015.07.013. dosing and enhance methane production via co-digestion in a bacterial-
Valverde-Pe
rez, B., 2015. Wastewater Resource Recovery via the Enhanced Biolog- microalgal process. Algal Res 20. https://doi.org/10.1016/j.algal.2016.10.010.
ical Phosphorus Removal and Recovery (EBP2R) Process Coupled with Green Wa
gner, D.S., Valverde-Pe
rez, B., Plo

sz, B.G., 2018. Light attenuation in photo-
Microalgae Cultivation. Technical University of Denmark, Kgs. Lyngby, bioreactors and algal pigmentation under different growth conditions e model
Denmark. identification and complexity assessment. Algal Res 35, 488e499. https://
Valverde-Pe
rez, Borja, Fuentes-Martínez, J.M., Flores-Alsina, X., Gernaey, K.V., doi.org/10.1016/j.algal.2018.08.019.
Huusom, J.K., Plo
sz, B.G., 2016a. Control structure design for resource recovery Wa
gner, D.S., Valverde-Pe
rez, B., Sæbø, M., Bregua de la Sotilla, M., Van Wagenen, J.,
using the enhanced biological phosphorus removal and recovery (EBP2R)
sz, B.G., 2016b. Towards a consensus-based biokinetic model for
Smets, B.F., Plo
activated sludge process. Chem. Eng. J. 296, 447e457. https://doi.org/10.1016/ green microalgae e. The ASM-A. Water Res. 103 https://doi.org/10.1016/
j.cej.2016.03.021. j.watres.2016.07.026.
Valverde-Pe
rez, B., Ramin, E., Smets, B.F., Plo

sz, B.G., 2015. EBP2R - an innovative Wang, L., Min, M., Li, Y., Chen, P., Chen, Y., Liu, Y., Wang, Y., Ruan, R., 2010. Cultivation
enhanced biological nutrient recovery activated sludge system to produce of green algae Chlorella sp. in different wastewaters from municipal wastewater
growth medium for green microalgae cultivation. Water Res. 68, 821e830. treatment plant. Appl. Biochem. Biotechnol. 162, 1174e1186. https://doi.org/
Valverde-Pe
rez, Borja, Wa
gner, D.S., Lo

ra

nt, B., Gülay, A., Smets, B.F., Plo

sz, B.G., 10.1007/s12010-009-8866-7.
2016b. Short-sludge age EBPR process e microbial and biochemical process Whitton, R., Le Me
vel, A., Pidou, M., Ometto, F., Villa, R., Jefferson, B., 2016. Influence
characterisation during reactor start-up and operation. Water Res. 104, of microalgal N and P composition on wastewater nutrient remediation. Water
320e329. https://doi.org/10.1016/j.watres.2016.08.026. Res. 91, 371e378.
Valverde-Pe
rez, B., Xing, W., Zachariae, A.A., Skadborg, M.M., Kjeldgaard, A.F., Winckelmann, D., Bleeke, F., Bergmann, P., Elle, C., Klo €ck, G., 2016. Detection of weed
Palomo, A., Smets, B.F., 2020. Cultivation of methanotrophic bacteria in a novel algae in open pond cultures of Cyanobacterium aponinum using PAM. Int.
bubble-free membrane bioreactor for microbial protein production. Bioresour. Aquat. Res. 8, 81e90.
Technol. 310, 123388. https://doi.org/10.1016/j.biortech.2020.123388.