Effective removal of the rotifer Brachionus calyciflorus from a Chlorella vulgaris microalgal culture by homogeneous solar photo-Fenton at neutral pH

In this study


Introduction
Microalgae are photosynthetic organisms which are widely spread in aquatic and terrestrial environments and have interesting marketable applications. For example, microalgae are used as food supplements in human nutrition or to produce valuable compounds, such as natural carotenoids, antioxidants, and vitamins (Singh et al., 2019;Vaz et al., 2016). Moreover, they are used in aquacultures to feed zooplankton, bivalve mollusks, and larval stages of crustaceans and some fish species (Brown et al., 1997;Sirakov et al., 2015). Recently, microalgae raised interest for their potential applications in biofuels, in biopolymer production, and even in pharmaceutical biotechnology (Yan et al., 2016;Zanchetta et al., 2021;Zewdie and Ali, 2020). Additionally, their demand as fertilizers makes them a lucrative market, with high expected growth over the next decade (Barsanti and Gualtieri, 2018;Uysal et al., 2015). Hence, there is accumulated interest for developing stable and functional growth protocols.
Despite their high utility, the mass production of microalgae is hindered by several factors, such as their high culturing costs and challenges associated with the use and up-scaling of cultivations systems, such as photobioreactors (PBRs) or open ponds (Borowitzka and Vonshak, 2017;Tredici et al., 2016). While the cultivation of microalgae in open ponds under natural irradiance is usually cheaper and more scalable than the cultivation in PBRs (Davis et al., 2011;Slade and Bauen, 2013), open systems are more vulnerable to biological contamination by grazers transmitted for instance through the air. Indeed, in natural ecosystems, microalgae are at the base of the food chain and they are predated by grazing microzooplankton, which has been documented as occasional contaminant of industrial cultures of microalgae (Carney et al., 2016;Molina et al., 2019;Yuan et al., 2018). Grazers thus represent a threat for the production and quality of the algal biomass and there is a need for effective and easy applicable treatments to inactivate the growth of these contaminants, particularly at an early culturing stage (Day et al., 2017).
Among grazers, rotifers, such as Brachionus calyciflorus, are small animals with a length between 100 and 200 µm (Nandini et al., 2007), which are voracious predators of microalgae (Rico-Martínez and Dodson, 1992). They thrive by producing diploid and haploid eggs, via amictic and mictic reproduction, respectively (Fussmann et al., 2007). Diploid eggs from amictic females will develop into females, while haploid eggs from mictic females will develop into males, or if fertilized, into resting eggs (cysts). Rotifers can consume a large variety of microalgae at high rates and their predation can cause a complete collapse of microalgae cultures (Abou-Shanab et al., 2016).
The cultures of microalgae are particularly vulnerable to rotifers when the cultivation is carried out at low cell densities (Deruyck et al., 2019b), hence contamination has to be detected at an early stage after culture inoculation and corrective measures have to be taken to avoid the failure of the microalgal cultures. Moreover, rotifers are abundant in various aquatic or moist environments such as lakes and rivers but also marshes, moist soils, mosses, tree holes, etc. (Gilbert, 2018;Phan et al., 2021;Walsh et al., 2014). Therefore, the contamination of microalgae cultures by rotifers may increase if natural freshwater is used. When the cultivation is carried out in open systems, wind and rain could also carry particles into the cultures, which potentially contain rotifers or their eggs/cysts.
Effective approaches for the removal of rotifers from microalgae cultures remain relatively unexplored, even if some innovative methods have been investigated and successfully applied so far, such as the use of hydrodynamic cavitation, surfactants, or botanical pesticides (Deruyck et al., 2019a;Huang et al., 2014aHuang et al., , 2014bKim et al., 2017;Zhang et al., 2021). Sodium hypochlorite (0.45 to 0.6 mg Cl L −1 ) was also successfully applied to inactivate rotifers in microalgae cultures, and to the best of our knowledge, this is the only study where a (plain) oxidative treatment was used for an in-situ inactivation of rotifers (Park et al., 2016). However, sodium hypochlorite and ozone treatments were used to produce bacteria-free cultures of rotifers, which shows that rotifers are equipped to face oxidative stress (Davis and Arnold, 1997;Douillet, 1998;Gonçalves and Gagnon, 2011). In our present study, we are aiming at the development of an effective oxidation process, which is fine-tuned to overcome the resistance of rotifers to oxidative damage and removes them from Chlorella vulgaris cultures while preserving the viability of the microalgae and allowing their regrowth after the decontamination treatment.
Advanced Oxidation Processes (AOPs) are a group of powerful and versatile treatments that have gathered significant interest within the water community over the last two decades . The common denominator between processes like the Fenton reaction, ozone (at neutral/basic pH) and UV/H 2 O 2 are the generation of very oxidative hydroxyl radicals (HO • ), second only to F − . While the Fenton process has the advantage of requiring simple reagents (Fe, H 2 O 2 ), it also has significant drawbacks, such as the operation at acidic pH or the generation of iron sludge (Pignatello et al., 2006). Nevertheless, recent advances in the field allow envisioning applications for matrices with neutral pH, by the use of chelators or sunlight (Ahile et al., 2020;Clarizia et al., 2017;O'Dowd and Pillai, 2020). The Fenton process has been used in (waste)water treatment to inactivate a wide array of microorganisms, such as bacteria (de la Obra Jiménez et al., 2020;Esteban García et al., 2018), viruses (Giannakis, 2018;Ortega-Gómez et al., 2015), fungi (Aguas et al., 2017), protozoa (Abeledo-Lameiro et al., 2019, even antibiotic resistant strains (Polo-López and Sánchez Pérez, 2021), and obtain purified effluents. It requires only Fe and H 2 O 2 , and is considered as an environmentally friendly method, which can be described by the following equations at near-neutral pH (Mosteo et al., 2020) In a recent study from our group , the application of the photo-Fenton (PF) process under (simulated) solar light with the help of a natural chelator (citrate) has effectively eliminated bacterial contaminants from a microalgae culture. Compared to existing oxidative decontaminating processes, this application has the advantage of performing well at neutral-basic pH thanks to citrate (Ruales-Lonfat et al., 2016), and utilizing solar light, a key component in algal cultures (Acién et al., 2012). Furthermore, despite the non-selective nature of hydroxyl radicals (Pignatello et al., 2006) and their capability of disrupting algae cells (Romero-Martínez et al., 2021;Torres et al., 2018), a selective action on bacteria was observed, leaving the microalgal culture relatively unharmed and with potential for post-treatment regrowth . As such, it would make an excellent case if a similar effect was attained in eliminating rotifers from a microalgal culture, without significantly hampering microalgae growth.
Considering the above, in the present study, two solar-based treatment methods, solar light/H 2 O 2 and PF, as well as corresponding control experiments, were applied on microalgae cultures that were experimentally contaminated with rotifers. In general, we address photoautotrophic culturing systems with our solar-driven decontamination process, because they are most commonly used for microalgae mass production and their commercial use (Udayan et al., 2022). In addition, our approach is also meant to be potentially applicable for closed photo-bioreactors that are increasingly used for the production of high value products and pharmaceuticals (Jha et al., 2017).
A further adaptation of our methodology to a mixotrophic growth mode in open plants or closed photobioreactors might be foreseeable. In mixotrophic cultivation microalgae simultaneously combine photosynthesis under light irradiation and organic carbon assimilation, which often results in higher microalgae biomass yields as compared to strict photoautotrophic cultivation (Zhan et al., 2017). The ultimate objective is to determine if the homogeneous citrate-modified PF method developed against bacterial contamination could also be used to remove a bigger and more complex organism that predates microalgae. To attain this goal, we studied i) the individual effects that the solar-based treatments exert on microalgae (Chlorella vulgaris) and rotifers (Brachionus calyciflorus) and ii) we tried to adapt the process parameters to completely eliminate rotifers from contaminated cultures while concomitantly achieve a good survival rate of microalgae. For this purpose, the effects on microalgae and rotifers were assessed during and immediately after different solar-based treatments as well as for up to two weeks later to assess long-term effects on microalgae growth and decontamination efficacy. Ultimately, the study aims to contribute to the development of a cheap and environmentally friendly in-situ oxidative treatment for the protection against biological contaminants in microalgae cultures.

Preparation of synthetic freshwater
A moderately hard synthetic freshwater was prepared with the following concentrations (mg L −1 ): 122.85 MgSO 4 ⋅7H 2 O, 60 CaSO 4 ⋅2H 2 O, 4 KCl, 96 NaHCO 3 in ultrapure water (milli-Q, Merck) (EPA, 2002). This solution was stored in the dark at ambient temperature and used as liquid medium for the cultivation of rotifers and the preparation of microalgae suspensions.

Cultivation of microalgae
Chlorella vulgaris (SAG 211-11b, Göttingen University, Germany) was cultivated to feed rotifers and to prepare standardized microalgae cultures, i.e., cultures with a dry weight of 100 mg L −1 and a cell density of ≈10 7 cell mL −1 , which were used to carry out experiments. The methods used to cultivate microalgae and to assess their viability, as well as the setup of glass-column PBRs, were largely identical to what was described in our previous study . Briefly, the setup of PBRs consists of glass columns (height: 50 cm, internal diameter 3.0 cm), a temperature controller, DULUX cool white lamps for illumination and mass flow controllers for CO 2 and air sparging. The following modifications were introduced in the methods employed in the present work: i) Tap water was replaced by synthetic freshwater as culture medium. ii) When microalgae were mixed with rotifers, samples to determine the cultivability of microalgae were filtered with a 40 µm mesh to remove rotifers prior to analysis. iii) Microalgae cultivation in glass-column PBRs was slightly altered: a) the nutrient concentration of the liquid mineral medium was doubled, b) instead of a gradually increasing light intensity, a constant intensity of 250 µmol m 2 s −1 of photosynthetically active radiation was used, c) the cultivation period was increased from 6 to 7 days, and d) liquid mineral medium was added to the culture after 3 days to increase the volume from 100 to 200 mL. iv) The microalgae used to carry out the experiments were always harvested after 7 days of growth and used fresh on the same day. About 100 mL of the microalgae culture was required to prepare the experiment and the stock to feed rotifers. Therefore, the remaining 100 mL were adjusted to 200 mL with liquid mineral medium if necessary, and the cultivation in a glass-column PBR was extended to 14 days. The microalgae harvested during this extended period were not used in experiments but to feed rotifers.

Preparation of microalgae stocks to feed rotifers
Synthetic freshwater was used for every step of the following procedure: microalgae were harvested, washed two times (4000 rcf for 3 min), and re-suspended in their initial volume. The cell count was determined, and microalgae were diluted to a cell density of 2⋅10 8 cells mL −1 . This stock was stored up to 4 days at 4 • C and used daily to feed rotifers.

Cell cultivability of microalgae on agar plates
The method is described in detail in our previous study , hence it is only briefly summarized here. A 1-mL sample was mixed with 30 µL of a catalase suspension (3 mg mL −1 ) to remove residual H 2 O 2 . Then, 0.1 mL of sample was then spread on agar plates prepared with mineral medium. When microalgae were mixed with rotifers, the sample was filtered with a 40 µm mesh before the addition of catalase. Agar plates were incubated at room temperature with a constant illumination of 60 µmol m −2 s −1 for 9 days. The number of cultivable cells of C. vulgaris was determined by manual counting.

Cultivation, harvesting and counting of rotifers
Brachionus calyciflorus was cultivated in synthetic freshwater in a 1-L culture and the population size was adjusted daily to 60 individuals mL −1 . The detailed protocols and methods to hatch eggs, to scale up and maintain the culture, and to enumerate the rotifers are described in the accompanying MethodsX article of the current submission (A comprehensive method for the stable and continuous production of Brachionus calyciflorus rotifers in a 1L culture using live microalgae feed).

Assessment of the rotifer viability during experiments
Viability of rotifers was determined at regular time points during the experiments by assessing rotifer growth in duplicate 1-mL samples, which were incubated up to 7 days (at 25 • C and under a low light intensity) in Petri dishes filled with 30 mL of synthetic water and microalgae (10 7 cells mL −1 ). Additionally, 30 µL of a 3-mg mL −1 bovine catalase solution (2000-5000 units mg −1 , Sigma-Aldrich) was added to neutralize the residual H 2 O 2 . Rotifers were counted every 24 h and monitoring was stopped one day after their concentration surpassed 100 rotifers per Petri dish (shown as a red line in figures, vide infra).

Practical experimental details of solar-based oxidative treatments
The decontamination reactors used in the solar-based treatments consisted of 250-mL beakers with floating magnetic stirring bars (DS6630, Thermo Scientific), both of which were acid-washed, rinsed, and autoclaved before carrying out the experiments. The setup of the solar simulator, light intensity (700 W m −2 ), and temperature (25 ± 0.5 • C) were the same as described in our previous study , but the stirring was decreased to 150 rpm. 200 mL of a microalgae and/or rotifers suspension was poured into each reactor and samples were retrieved to assess the viability of microalgae and rotifers at t=0. Reagents were then added to the reactors, according to the treatment, in the following order: 1) citric acid, 2) Fe 2+ , and 3) H 2 O 2 , with two-minute intervals in between. The lamp was switched on and the experiment was carried out for 2 h. The first sample for the analyses of dissolved Fe 2+ /Fe 3+ and H 2 O 2 was taken 30 s after the addition of reagents.
Samples were taken regularly to determine the viability of rotifers and microalgae, concentrations of dissolved Fe 2+ /Fe 3+ and H 2 O 2 , and pH. For all the rotifer experiments, the average pH value at t=0 was 8.0 ± 0.1 for synthetic freshwater alone and 7.6 ± 0.1 when microalgae were added; it dropped to 6.9 ± 0.1 in both cases when Fe 2+ was added. However, the average pH value increased during experiments to reach 7.4 ± 0.1 in synthetic freshwater and 7.7 ± 0.2 in microalgae cultures at the end of the 2-h treatment.
If a treated culture was cultivated after the 2-h treatment, H 2 O 2 was neutralized in the reactor by addition of 1.4 mL of a bovine catalase solution (3 mg mL −1 ). Then, the stirring bar was removed, and nutrients were added to reach 1/4 of the concentration of the liquid mineral medium. The reactor was placed under a lamp (Easy LED Universal Full Spectrum, Aquatlantis) with a light intensity of 100 µmol m −2 s −1 in a water bath at 25 ± 1.5 • C. A mass flow controller provided to the culture a constant air flow of 0.1 L N min −1 (2 % CO 2 (v/v)), and 200 µL of a 1 M NaOH solution were added after 30 min to stabilize the pH value around 7. The algal cell and rotifer concentrations were determined daily by counting under a light microscope (in duplicate) or a zoom stereomicroscope (in triplicate), respectively. Water evaporation was compensated daily by addition of synthetic freshwater. The stability of the pH value was verified every three days.

General experimental outline of solar-based oxidative treatments
Four different sets of experiments were carried out in microalgae cultures or synthetic freshwater with different treatments and concentrations of rotifers, always at a simulated solar light intensity of 700 W/ m 2 . The first set of experiments focused on microalgae cultures (dry weight: 100 mg L −1 , cell density ≈10 7 cells mL −1 ) without rotifers and aimed to assess the effect of i) simulated solar light alone, ii) solar light and Fe 2+ with citric acid (1 mg L −1 Fe 2+ , 17.5 mg L −1 citric acid), iii) solar light and H 2 O 2 (20 mg L −1 H 2 O 2 ), and iv) the homogeneous citratemodified PF (1 mg L −1 Fe 2+ , 20 mg L −1 H 2 O 2 , 17.5 mg L −1 citric acid). Further sets of experiments consisted of the same treatments but used an inoculum of 20 rotifers per mL, in synthetic freshwater without microalgae (second set) and in a microalgae culture (third set), respectively. The fourth set of experiments consisted of 5 rotifers mL −1 in a microalgae culture to which a citrate-modified PF treatment (1 mg L −1 Fe 2+ , 20 mg L −1 H 2 O 2 , 17.5 mg L −1 citric acid) was applied. Two biological replicates were carried out for each experimental condition.
Microalgae cultures with 20 rotifers mL −1 were recultivated after their treatment with solar light/H 2 O 2 and citrate-modified PF. Microalgae cultures with 5 rotifers mL −1 were also recultivated after the treatment with citrate-modified PF. Additionally, microalgae cultures with and without rotifers, which were not treated, were cultivated under the same conditions and their data were used as a reference.

H 2 O 2 and dissolved iron measurements
1 mL of a filtered sample was mixed with 10-20 µL of titanium (IV) oxysulfate to measure the concentration of H 2 O 2 by colorimetry (DIN • 38.402 H15) at 410 nm with a spectrophotometer (UV-1800, Shimadzu). The concentration of H 2 O 2 was determined with a calibration curve of seven points between 0 and 25 mg mL −1 (R 2 > 0.99). The Ferrozine method was used to measure spectrophotometrically at 562 nm the dissolved Fe 2+ and Fe 3+ (Viollier et al., 2000). 1.6 mL of a filtered sample was mixed with 0.2 mL of a 4.9 mM ferrozine solution, 0.2 mL of hydroxylamine hydrochloride solution (10 % w/w) and 0.5 mL of an acetate buffer (pH 4.66). A calibration curve of six points between 0 and 2 mg mL −1 was used as a reference to determine the concentration of dissolved Fe 2+ /Fe 3+ and the limit of quantification (LOQ) was 0.01 mg L −1 . Filtrations were carried out with syringe filters (0.45 µm). Error bars represent the standard error from the mean.

Data treatment
Data on the viability of microalgae, from the plating on agar of samples collected during the treatment time, were standardized to the initial concentration and are given as colony forming units (CFU), which was expressed graphically as the inactivation of cells in logarithmic units (logU) per mL (logCFU/mL). Data on algal cell counts, from the enumeration under a light microscope of samples collected during the post-treatment recultivation, were also standardized to the initial concentration and are given as the concentration of cells per mL in logU (log cells/mL). Rotifer counts are given as the number of swimming female rotifers per Petri dish or per mL. Errors bars represent the standard error of the mean. Swimming female rotifers were the focus of this study, and therefore, eggs, male rotifers and motile/inactive female rotifers are not shown in the results. Male rotifers were excluded from the study because they accounted for only 10 to 20 % of the population, thanks to the precise daily maintenance routine. Moreover, male rotifers do not produce eggs, hence they do not contribute to population regrowth after treatment, and had a neglectable weight, hence their contribution to consumption of reagents was negligible.
The limits of detection for the microalgae cultivability on agar plates and cell enumeration under the light microscope were 10 CFU mL −1 and 10 4 cells mL −1 , respectively. The limits of detection for the enumeration of rotifers were 1 unit (rotifer) per Petri dish for viability tests carried out during the treatments, and 1 rotifer mL −1 for the monitoring carried out during the cultivation of treated cultures.

Moderate inhibitory effects of solar-based oxidative processes on microalgae cultures
First experiments were carried out with cultures of C. vulgaris without addition of rotifers to assess inhibitory effects of two different solar-based processes (solar light/H 2 O 2 , and homogeneous citratemodified photo-Fenton) and their controls (solar light alone, solar light/Fe 2+ with citric acid) on the cultivability of algal cells during the treatment and to determine the associated consumption of reagents (Fig. 1). The results show that the exposure to solar light alone had not detrimental effect on microalgae viability and neither had the addition of Fe 2+ and citric acid. Indeed, iron is an important nutrient that plays a key role in many metabolic processes of microalgae, such as photosynthesis, respiration and nitrogen uptake (Marchetti and Maldonado, 2016). Moreover, the concentration of iron used in this study was in the range, or below, the concentration usually used to cultivate C. vulgaris (Doušková et al., 2010;Golub et al., 2018). Therefore, a solar light intensity of 700 W m −2 , which is in the range of ambient irradiation, and the presence of a low concentration of iron, was unlikely to have a significant negative impact on the microalgae viability. However, the solar light/H 2 O 2 and citrate-modified PF treatments inactivated 2.46 ± 0.13 and 1.83 ± 0.31 logU of microalgae per mL of culture after 2 h, respectively (Fig. 1A). These results corroborate observations made in our previous study , i.e., that the solar light/H 2 O 2 treatment is more efficient than the citrate-modified PF in inactivating the growth of microalgae. However, in the present study, the loss of microalgae was moderate for both treatments due to an adapted H 2 O 2 concentration, which was decreased to 20 mg L −1 compared to our previous work. Indeed, a residual, post-treatment concentration of viable microalgae in the range of 10 5 cells mL −1 is expected to be high enough to enable successful regrowth of the culture (Bohutskyi et al., 2016).
The consumption curves of reagents during treatment of microalgae (Fig. 1B) were compared with curves of the same treatments carried out in absence of microalgae in ultrapure water and synthetic freshwater ( Supplementary Fig. S1). The results show that microalgae induce the high consumption of H 2 O 2 during solar light/H 2 O 2 and citrate-modified PF treatments (i.e., this is not a matrix effect). The consumption of H 2 O 2 in the PF reaction was 6 or 8 mg L −1 when the reaction was carried out in microalgae cultures or synthetic water, respectively. Most likely, H 2 O 2 is initially consumed by reaction with Fe 2+ , followed by reaction with algal cells. In terms of iron, it is also likely that a fraction of the dissolved Fe 2+ /Fe 3+ was consumed by interaction with microalgae (binding or uptake). However, the iron concentration of the culture medium was also affected using freshwater, which is most likely due to the nearneutral pH value that accelerates its precipitation. Our data show that the citrate-modified homogeneous PF reaction occurred mainly during the first 30 min, most probably followed by a considerably milder heterogeneous reaction (in terms of HO • generation), driven by the iron precipitates. The photo-Fenton process following this step should be slower due to the heterogeneous nature of precipitated iron (iron oxides).

Inhibitory effects of different solar-based processes on the regrowth of rotifer cultures in the absence of microalgae
The same solar-based processes as detailed above were applied to a rotifer culture with 20 individuals mL −1 in synthetic freshwater, to assess the effects of the different treatments on their viability and to determine the consumption of reagents during the treatment period (Fig. 2). Rotifers were neither adversely impacted by solar light in the range of ambient irradiation ( Fig. 2A), nor by a concentration of 1 mg L −1 of Fe 2+ and 17.5 mg L −1 of citric acid (Supplementary Fig. S2). Indeed, adverse effects of iron on the rotifer growth were previously only reported for a much higher concentration (Han et al., 2021). Therefore, the rotifer inactivation observed during the solar light/H 2 O 2 and citrate-modified PF (Fig. 2B, 2C) resulted predominantly from oxidation of rotifers by the solar/H 2 O 2 process or the generated radicals, such as HO • , produced during the photo-Fenton reaction ( Eqs. 1-4).
The results in Fig. 2B show that rotifers were fully inactivated by the solar light/H 2 O 2 treatment within 90 min. Moreover, the regrowth was already strongly reduced after 60 min of treatment with an average value of only 11 ± 11 rotifers per Petri dish after 7 days of incubation. As far as the efficacy of the PF process is concerned, rotifers were fully  inactivated by the citrate-modified PF treatment in 120 min and their regrowth was already severely impacted after 90 min of treatment, as an average of only 42 ± 42 rotifers were counted in Petri dishes after 7 days of incubation (Fig. 2C). Comparatively, the growth rates of untreated rotifers (i.e., samples taken at t=0) were much higher and the incubation was stopped after 3 days since the average value reached 196 ± 12 and 277 ± 95 rotifers.
These results show that rotifers can be fully inactivated, and their subsequent regrowth prevented by applying certain oxidative treatments (i.e., H 2 O 2 and citrate modified PF). The solar light/H 2 O 2 treatment appeared to be the main driving force of the rotifer inactivation, being more efficient than the citrate-modified PF in yielding more pronounced inactivation, as it has also been observed in the treatment of microalgae alone, and quicker regrowth inhibition. Rotifers are bigger in size than microalgae, and, therefore, it is possible that they are also more resistant to a strong, but short, external oxidative stress burst induced by the citrate-modified PF reaction. However, the efficacy difference observed between the two treatments remains relatively narrow (only one sampling time-step difference; 90 vs. 120 min in treatment, 60 vs. 90-min effect in regrowth).
A population of 20 rotifers mL −1 did not consume much of H 2 O 2 and dissolved Fe 2+ /Fe 3+ when compared to the corresponding curves obtained for microalgae cultures (Fig. 2D). This is likely due to the comparatively lower total biomass and contact surface in rotifer cultures. Dry weight of a single rotifer is estimated between 0.11 and 0.47 µg (Dumont et al., 1975), which results in 2.2 to 9.4 mg L −1 for a concentration of 20 rotifers mL −1 . Dry weight of microalgae was 100 mg L −1 for a higher number of organisms (≈10 7 cells mL −1 ), hence the specific surface for reagent interaction was higher for microalgae cultures. These results indicate that the inactivation of rotifers may be feasible in a microalgae culture without leading to a faster reagent consumption than observed for cultures of microalgae alone, and therefore, without potentially impacting operation and cost-related factors.

Treatment of highly contaminated microalgae cultures with solar light/H 2 O 2 and citrate-modified photo-Fenton processes, and subsequent regrowth of treated cultures
Microalgae cultures were contaminated with a rather high starting concentration of 20 rotifers mL −1 and then treated with solar light/H 2 O 2 or citrate-modified PF treatments. Solar light and solar light/Fe 2+ (with citric acid) processes were carried out as controls. In all experiments, the cultivability of microalgae, regrowth of rotifers, and concentrations of H 2 O 2 and dissolved Fe 2+ /Fe 3+ were monitored during the treatment ( Fig. 3 and Supplementary Fig. S3).
Microalgae were inactivated by 1.8 ± 0.13 and 2.00 ± 0.01 logU per mL of culture by the citrate-modified PF and solar light/H 2 O 2 treatments, respectively (Fig. 3A), which was similar to the results obtained with microalgae cultures that were not contaminated with rotifers. However, the inactivation of rotifers upon treatment with citratemodified PF (Fig. 3B) and solar light/ H 2 O 2 (Fig. 3C) was lower when compared to their treatment in the absence of microalgae. As a result, the rotifer population recovered during the 7-day incubation, even when cultures were treated for 2 h (contrary to their full inactivation in absence of microalgae, Fig. 2B, 2C). This indicates that treatment efficacy was mainly impacted by the presence of microalgae, because they dominated the consumption of H 2 O 2 (Fig. 3D), and therefore, decreased its oxidative effect on rotifers. However, recovery of rotifers was carried out in Petri dishes under ideal culturing conditions adapted for rotifers and results may differ under conditions ideal for microalgae growth, i.e., in liquid mineral medium, illuminated, and bubbled with an air/CO 2 flow. Hence, the results obtained here represent a worst-case scenario for a field-application.
After treatment with oxidative solar-based processes, the treated cultures were cultivated for 7 days and the ability of microalgae and rotifers to recover from the different treatments was assessed by the monitoring of algal cell and rotifer concentrations (Fig. 4). Regrowth was compared to microalgae cultures with and without rotifer contamination (20 rotifers mL −1 ) that were not treated. Regrowth of rotifers was, on average, not prevented by the treatments, but only delayed, indicating that not all rotifers were killed during the treatment. Indeed, rotifers reached moderate concentrations of 9 ± 8 and 16 ± 16 rotifers mL −1 after 7 days of cultivation in cultures treated with solar light/H 2 O 2 and citrate-modified PF, respectively. However, these concentrations were much lower than the concentration of rotifers of untreated contaminated microalgae cultures, where rotifers reached an average concentration of 145 ± 26 individuals mL −1 .
While microalgae cultures recovered from both solar-based treatments, a negative effect is visible. Cell concentration decreased during the first four days, most probably due to the residual effect of the inflicted oxidative damages, and then increased until the end of the cultivation. Finally, cultures reached cell density of 7.03 ± 0.06 and 7.70 ± 0.05 logU mL −1 for the solar light/H 2 O 2 and citrate-modified PF treatments after 7 days, respectively, which was lower than the 8.46 ± 0.02 logU mL −1 reached by untreated cultures without rotifers, in 5 days. Microalgae cultures showed better regrowth when they were treated with citrate-modified PF instead of solar light/H 2 O 2 , which is consistent with the results of their cultivability on agar plates.
In comparison to untreated microalgae cultures, with and without rotifers, we demonstrate the considerable impact that rotifers have on algal growth. Cell densities of contaminated cultures fluctuated between 6.36 and 7.09 logU mL −1 , which was clearly below those of cultures that were not contaminated with rotifers (8.46 LogU mL −1 ). Additionally, the increasing population of rotifers (145 rotifers mL −1 after 7 days) indicated that their impact was likely increasing and that microalgae were most probably destined to be consumed.

Treatment of low-contaminated microalgae cultures with citratemodified photo-Fenton
Our results indicate that the intensity of the oxidative treatments is appropriate to largely preserve the viability of C. vulgaris cultures. It was, however, not sufficient to fully inactivate a contamination of 20 rotifers per mL −1 in a microalgae culture. Therefore, the citratemodified PF reaction, which affected microalgae less than the solar/ H 2 O 2 process, was applied to a lower concentration of only 5 rotifers mL −1 . This corresponds to a measure taken at an earlier stage of contamination in a field application (Yuan et al., 2018).
The inactivation of the microalgae and rotifers, and the concentration of reagents were monitored to assess the impact of the PF treatment (Fig. 5). The intensity of the treatment is modulated by the biomass of microalgae: their inactivation and the consumption of reagents were similar to the results obtained previously for microalgae cultures with and without rotifers. Microalgae were moderately affected and most of the reagents were consumed at the end of the 2-h treatment (Fig. 5A). Rotifers, however, were fully inactivated by the citrate-modified PF treatment in 90 min (Fig. 5B). This indicates that a rotifer contamination that is detected at an earlier stage can be successfully treated.
Treated cultures were cultivated for 14 days to confirm that the inactivation of rotifers was permanent and to monitor the regrowth of microalgae (Fig. 6). The successful inactivation of rotifers was confirmed, as no regrowth was observed during the 14 days of posttreatment cultivation. Treatment delayed growth of microalgae by two days only: microalgae of treated cultures were able to reach their stationary phase after only nine days of cultivation (8.23 ± 0.06 logU mL −1 ) while non-contaminated microalgae cultures reached their stationary phase (8.27 ± 0.03 logU mL −1 ) in seven days.
This short delay in growth after treatment is a modest drawback compared to an untreated culture, as complete restart of the cultivation likely takes considerably longer. The threat of rotifers, even at a low concentration of 5 individuals per mL, was demonstrated in untreated, but contaminated, microalgae cultures. Microalgae were first moderately impacted by rotifers during the first eight days and their cell density approached the value of the uncontaminated culture by reaching up to 8.07 ± 0.1 logU mL −1 . However, during the second week of cultivation, rotifer numbers reached 1521 ± 412 rotifers mL −1 and caused cultures to fail. This demonstrates the efficacy of the citratemodified PF treatment to stop rotifer contamination in a microalgae culture, but also the dramatic decay of microalgae and strong propagation of rotifers, if no countermeasures are applied.
With the citrate-modified PF treatment (1 mg Fe 2+ , 20 mg L −1 H 2 O 2 , 17.5 mg L −1 citric acid), a rotifer contamination was successfully removed from a microalgae culture, while microalgae were able to regrow. The performance achieved was comparable to other known chemical methods such as botanical pesticides, surfactants, or disinfectants, but without the associated disinfection by-products (Farinelli et al., 2021). For example, a complete inactivation of 5 rotifers mL −1 was achieved with 3 µM of cetyltrimethylammonium bromide in C. vulgaris cultures (biomass dry weight: 160 mg L −1 ) (Deruyck et al., 2019a), and with 1.8 µg L −1 of toosendanin in Chlorella sp. cultures (cell density 2⋅10 −7 cells mL −1 ) (Huang et al., 2014a). However, the citrate-modified PF process has the additional advantage of not being persistent in the water after the treatment. Sodium hypochlorite (0.45 to 0.6 mg Cl L −1 ) supplied every two hours during 1 to 3 days inactivated a concentration of 7 rotifers mL −1 in Chlorella kessleri cultures (biomass dry weight: 60 to 90 mg L −1 ) (Park et al., 2016). However, this treatment time was much longer than for the citrate-modified PF process, which, in our conditions, required only 90 min. To summarize, the homogeneous, citrate-modified solar PF process presents a promising alternative treatment method that should be evaluated further.

Conclusions
A citrate-modified photo-Fenton process was used to treat microalgae cultures (C. vulgaris) that were artificially contaminated with rotifers (B. calyciflorus). The time-controlled treatment aimed to reach a full inactivation of rotifers and their eggs while minimizing harm to microalgae and, thus, allowing their continued cultivation. The treatment was effectively applied to a concentration of 5 rotifers mL −1 in the microalgae culture, corresponding to an intervention occurring at an early contamination stage.
In cultures containing both rotifers and microalgae, rotifers were fully inactivated after a short (90 min) oxidative treatment and no regrowth was observed during the following 14 days of cultivation. In contrast to rotifers, microalgae recovered from the oxidative stress and reached stationary phase with similar cell densities as obtained for noncontaminated microalgae cultures with a delay of two days.
Both the citrate-modified photo-Fenton and solar light/H 2 O 2 processes have similar efficacies for the inactivation of rotifers in contaminated microalgae cultures. Nevertheless, the citrate-modified photo-Fenton tends to inflict less damage on C. vulgaris and offers a better control on the reagent removal as more of the H 2 O 2 and all dissolved Fe 2+ /Fe 3+ were consumed at the end of a 2-h treatment.
The study shows the importance of applying treatment at an early stage of contamination. Indeed, citrate-modified photo-Fenton and solar light/H 2 O 2 treatments failed to fully inactivate rotifers in microalgae cultures at a concentration of 20 rotifers mL −1 . The results demonstrate the successful application of an oxidative treatment for the in-situ inactivation of microalgae predators in an environmentally friendly way. The next step of the project would be to apply the treatment at larger scale in a commercial PBR designed to produce microalgae, and to continue assessment of its potential against other biological (e.g., ciliates, fungi, or invading microalgae) and chemical contaminants of microalgae cultures (Yang et al., 2021;Zhu et al., 2020).

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data Availability
Data will be made available on request.

Acknowledgments
This work was carried out within the framework of the European COST Action ES1408 (EUALGAE) supported by the Swiss National Science Foundation. Stefanos Giannakis acknowledges the Spanish Ministry of Science, Innovation and Universities (MICIU) for the Ramón y Cajal Fellowship (RYC2018-024033-I). The Swiss Competence Center for Bioenergy Research (SCCER BIOSWEET) and their partners are acknowledged for valuable discussions and financial support in the early stage of the project.

Supplementary materials
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.watres.2022.119301. [Microalgae] 0 = 10 7 cells mL −1 . Continuous and dash lines between points are only used as a visual help.