Anaerobic Co-Digestion Effluent as Substrate for Chlorella vulgaris and Scenedesmus obliquus Cultivation

: Anaerobic digestate supernatant can be used as a nutrient source for microalgae cultivation, thus integrating phytoremediation processes with high value products storage in microalgae biomass. Microalgae are able to use nitrogen and phosphorous from digestate, but high nutrient concentration can cause growth inhibition. In this study, two microalgae strains ( C. vulgaris and S. obliquus ) were cultivated on the anaerobic co-digestion supernatant (obtained from the organic fraction of municipal solid waste (OFMSW) and waste activated sludge (WAS)) in a preliminary Petri plate screening at di ﬀ erent dilutions (1:10 and 1:5) using a synthetic medium (ISO) and tap water (TW). Direct Nile red screening was applied on colonies to preliminarily identify hydrophobic compound storage and then a batch test was performed (without air insu ﬄ ation). Results show that C. vulgaris was able to grow on digestate supernatant 1:5 diluted, while Nile red screening allowed the preliminary detection of hydrophobic compound storage in colonies. The analysis carried out at the end of the test on ammonia, phosphate, nitrate and sulphate showed a removal percentage of 47.5 ± 0.8%, 65.0 ± 6.0%, 95.0 ± 3.0% and 99.5 ± 0.1%, respectively.


Introduction
The disposal of anaerobic supernatants can lead to several environmental problems, mainly due to their high nitrogen content. Their direct spreading onto agricultural land can cause nutrient emissions and cause eutrophication of nearby surface and ground waters [1][2][3] and NH 3 volatilization. Moreover, supernatants from digester-treated municipal organic waste and sewage sludge are likely to contain pathogens, heavy metals and microplastics (phthalates contamination) [4,5]. A strategy to mitigate the environmental impact of digestate or to avoid nutrient recirculation within the wastewater treatment plant is to use supernatants as a nutrient source for aquatic microalgae cultivation [6]. Additionally, microalgae cultivation could be coupled with lipid intracellular accumulation for subsequent biodiesel production since microalgae have a short carbon period for lipid storage and a faster photosynthetic efficiency (3-8%) compared to plant oils (0.5%) [7].
Microalgae are unicellular microorganisms, with photosynthetic activity and the capacity to grow on a wide variety of nutrient sources, including anaerobic digestion (AD) supernatant [8][9][10]. The AD effluent composition (i.e., high nitrogen and phosphorus concentration) has positive effects Table 1. Organic fraction of municipal solid waste and waste activated sludge (OFMSW-WAS) digestate characterization (n = 2).

Microalgae Strains, Maintenance and Analysis
Scenedesmus obliquus and Chlorella vulgaris (wild type) were provided by Federico II University of Naples (Naples, Italy) and maintained in synthetic medium (ISO 8692) [32]. Optical density (OD 680 and OD 750 nm), cellular count and dry weight were carried out to monitor cellular growth. Optical density was measured with a spectrophotometer UV4 100 Heλos Υ. Cellular count was monitored with a Bürke chamber on a Lika microscope using a 10 µL aliquot. Dry weight was measured as total solid (TS) method (Equation (1)), while growth rate was calculated using Equation (2) [33]. All tests and analysis were carried out in duplicate.
Dry weight g l −1 = g f2 − g f1 /l sample (1) where g f2 is the weight of filter with sample after 2 days at 105 • C, g f1 is the filter weight.

Petri Plate and Nile Red Qualitative Test
Nile red coloring has been used as an alternative method to identify hydrophobic compound storage in several kinds of biomass, including animal cells, bacteria and microalgae [34][35][36]. This technique allows for qualitative and quantitative analysis and it represents a valid and alternative simple methodology for lipid determination compared to conventional solvent-based extraction methods [37]. For example, classic chemical methodology for lipid quantification uses thin-layer chromatography (TLC), gas chromatography (GC) or high-performance liquid chromatography (HPLC) associated with complete lipid extraction with solvents that require labor-and time-intensive activity that would not allow for easy large sample screening. Therefore, Nile red coloring stands as an ideal method for scale up screening [38]. Nile red drop coloring was applied on a Petri plate screening on microalgae colonies, where S. obliquus and C. vulgaris were tested on solid media (Agar agar 1.5%, Sigma Aldrich, St. Louis, MO, USA) on Petri plates. According to the results of Zuliani at al. [9], Uggetti at al. [20] and Cho et al. [22], in this preliminary study, liquid digestate was diluted with tap water (DIG+TW) and ISO 8692 medium (DIG+ISO) with two different dilutions (1:10 and 1:5). OFMSW-WAS digestate's NH 4 + concentration (573 ± 124 mg L −1 ) could inhibit microalgae proliferation; typically an ammonium concentration higher than 160 mg L −1 has a toxic effect on microalgae biomass growth [20,23], so strong dilution is required to have microalgae growth. Levels of 2.10 klux and 3.9 klux light irradiation were used to observe high and low illumination effects on microalgae growth in the selected media. After 8 days, Nile red reagent (Sigma Aldrich) was dropped on colonies to observe hydrophobic compounds accumulating in the microalgae biomass. Nile red drop coloration was applied, based on the Greenspan et al. publication [34], where a Nile red solution (1000 µg mL −1 ) was diluted 1:100 with acetone (Sigma Aldrich) and used for qualitative analysis on microalgae strains. Onto every colony was poured two Nile red solution drops; after 20 min in dark conditions the Petri plate was arranged on an ultraviolet plate to look at the Nile red fluorescence. All tests and analysis were performed in duplicate.

Microalgae Growth: Batch Experiment Set-Up
S. obliquus and C. vulgaris growth was tested in mixotrophic conditions on liquid OFMSW-WAS supernatant diluted by 1:5 with ISO 8692 or tap water (named DIG+ISO and DIG+TW respectively). Phototrophic growth conditions (ISO and TW) were used as control and all test conditions were carried out in duplicate. Inoculum concentration was 3.8 ± 0.5 million cells mL −1 and 6.0 ± 1.0 million cells mL −1 for S. obliquus and C. vulgaris, respectively. Mechanical agitation (10 rcf) and light irradiation (3.9 klux) were continuous, no air bubbling was added to have a more realistic microalgae growth and nitrogen uptake as well as to minimize ammonia removal by stripping. The test duration was 8 days and experiments were performed in 300 mL Kimax flasks at a stable temperature of 20 • C ± 1 with Energies 2020, 13, 4880 4 of 12 initial inoculum diluted 1:5. Chlorophyll a, b and carotenoids accumulation in microalgae biomass were measured at the end of each test [39]. Anionic and cationic chromatography analyses were performed to show anionic and cationic ion variation before (Table 2) and after microalgal treatment on OFMSW-WAS digestate. All test conditions and analyses were performed in duplicate. Table 2. Characterization of initial cation and anion concentration in synthetic medium (ISO), tap water (TW), digestate diluted 1:5 with ISO and tap water (DIG+ISO and DIG+TW) before microalgae inoculum (n = 2).

Analytical Method
Tests for ammonia nitrogen, alkalinity, total and volatile solid (TS, VS), volatile fatty acids (VFA), pH and soluble chemical oxygen demand (sCOD) were performed according to APHA, APAT AWWA WET methods and ISA_CNR methodology [40,41]. Chlorophyll a, b and carotenoids analysis were carried out with methanol extraction (5 mL for 2.5 mL of microalgae sample) [42] for 1 h at 60 • C. After centrifugation at 1290 rcf the supernatant obtained (methanol plus chlorophylls and carotenoids) was measured with a spectrophotometer at λ 665.2 nm, 652.4 nm, 470 nm [43,44]. Chlorophyll a, b (Ch a and Ch b) and carotenoids quantification were obtained with Equations (3)-(5). The chlorophylls/carotenoids ratio analysis could be used as a marker to identify cultivation stress conditions [9,45].

Petri Plate: Nile Red Coloring Test
Petri plate screening showed that S. obliquus cultivated on both digestates (DIG+ISO and DIG+TW), at both dilutions (1:5 and 1:10) and under both irradiations (2.1 klux and 3.9 klux) was inhibited, while C. vulgaris showed dense green colonies (Figure 1). S. obliquus inhibition of growth was also reported by Zuliani et al. [9], however they also reported inhibition for the growth of C. vulgaris. After microalgae colony formation, Nile red drop coloring was used to identify hydrophobic compounds stored inside the microalgae. Control colonies were not colored, and they were exposed on ultraviolet as blank. Figure 2 shows fluorescent emission by colonies after Nile red coloring: C. vulgaris cultivated on DIG+TW 1:5 showed the highest visible emissions, other colonies did not show significant emissions when compared to blanks.

Petri Plate: Nile Red Coloring Test
Petri plate screening showed that S. obliquus cultivated on both digestates (DIG+ISO and DIG+TW), at both dilutions (1:5 and 1:10) and under both irradiations (2.1 klux and 3.9 klux) was inhibited, while C. vulgaris showed dense green colonies (Figure 1). S. obliquus inhibition of growth was also reported by Zuliani et al. [9], however they also reported inhibition for the growth of C. vulgaris. After microalgae colony formation, Nile red drop coloring was used to identify hydrophobic compounds stored inside the microalgae. Control colonies were not colored, and they were exposed on ultraviolet as blank. Figure 2 shows fluorescent emission by colonies after Nile red coloring: C. vulgaris cultivated on DIG+TW 1:5 showed the highest visible emissions, other colonies did not show significant emissions when compared to blanks.

Petri Plate: Nile Red Coloring Test
Petri plate screening showed that S. obliquus cultivated on both digestates (DIG+ISO and DIG+TW), at both dilutions (1:5 and 1:10) and under both irradiations (2.1 klux and 3.9 klux) was inhibited, while C. vulgaris showed dense green colonies (Figure 1). S. obliquus inhibition of growth was also reported by Zuliani et al. [9], however they also reported inhibition for the growth of C. vulgaris. After microalgae colony formation, Nile red drop coloring was used to identify hydrophobic compounds stored inside the microalgae. Control colonies were not colored, and they were exposed on ultraviolet as blank. Figure 2 shows fluorescent emission by colonies after Nile red coloring: C. vulgaris cultivated on DIG+TW 1:5 showed the highest visible emissions, other colonies did not show significant emissions when compared to blanks.    S. obliquus and C. vulgaris batch tests were carried out with OFMSW digestate diluted 1:5 with ISO 8692 (DIG+ISO) and tap water (DIG+TW). Control conditions were ISO 8692 (ISO) only and tap water (TW) only in mixotrophic growth conditions. C. vulgaris and S. obliquus cell count and OD 680 showed C. vulgaris had higher growth in both media than S. obliquus (Table 3). These results confirmed the data previously obtained in Petri plate screening, where S. obliquus growth was inhibited (Figure 2). Similar results were obtained by Zuliani et al. [9], where C. vulgaris and S. obliquus were cultivated on filtered and autoclaved diluted digestate. In Zuliani et al. [9], C. vulgaris and S. obliquus growth were 55 and 15 million cells mL −1 respectively. In this study, C. vulgaris and S. obliquus OD 680 results were lower than those obtained by Zuliani et al. [9] (2.0 and 1.8 times lower for C. vulgaris and S. obliquus, respectively). The difference could be associated with the experimental set-up difference, for example, in this study digestate was not sterilize and air bubbling was not applied in the batch test. Additionally, visual observation using an optical microscope (40×) showed a significant morphological change in C. vulgaris cells at the end of test, where C. vulgaris cells looked bigger than the initial cells. Probably this morphological change was associated with the accumulation of secondary high-value product in the cell [47]. C. vulgaris and S. obliquus dry weight at the end of the experiment (Table 3) showed similar biomass production in DIG+ISO and DIG+TW. These values are lower than those reported by Zuliani et al. [9], who obtained 2.00 g L −1 and 1.75 g L −1 for C. vulgaris and S. obliquus, respectively, but this could be due to medium sterilization and air insufflation that can help biomass growth [26]. Other studies reported a C. vulgaris biomass production of 0.6-0.7 g L −1 on diluted OFMSW digestate (1:10) after filtration at 0.2 µm or preautoclaved [22,23]. Bjornsson et al. [48] using Scenedesmus sp. on swine manure digestate diluted with tap water observed a biomass production of 0.55 g L −1 and NH 3 and PO 4 3− total removal. Bjornsson et al. [48] suggested that some fresh water micro-and macronutrients could be absent in artificial medium or digestate and this did not permit microalgae growth. For example, Energies 2020, 13, 4880 7 of 12 Mg 2+ is essential for microalgae growth. Mg 2+ ion is an essential constituent of chlorophyll and its concentration in tap water could be high enough to facilitate microalgae proliferation.
Growth rate (µ) values (Table 3) showed that the growth rate of C. vulgaris was faster than the growth rate of S. obliquus in both DIG+ISO and DIG+TW. Values obtained for C. vulgaris were comparable with the ISO control condition. In addition, C. vulgaris showed a longer exponential growth phase compared to S. obliquus (4 days and 2 days, respectively). As reported by Massa et al. [49] who had tested S. obliquus on OFMSW digestate diluted 1:3 in synthetic medium, this strain showed no growth in the presence of a low digestate dilution as obtained in this study. Uggetti et al. [20] showed a S. obliquus growth rate between 0.04 d −1 and 0.90 d −1 on OFMSW digestate diluted 1:5 and 1:10, respectively. Veronesiv et al. [50] tested Chlorella sp. on agro-zootechnical ultrafiltered digestate 1:10 diluted and they obtained a growth rate of 0.2 ± 0.01 d −1 . Growth rates obtained in this study were comparable to those reported by these publications where digestate was autoclaved and filtered before use and microalgae cultivation was performed with air bubbling application.

Nutrient Removal and Chlorophyll a, b and Carotenoid Accumulation
Petri plates test results encouraged batch testing with a supernatant diluted 1:5, with 199 ± 9 mg L −1 initial ammonia concentration in both test conditions: DIG+ISO and DIG+TW. As reported in Figure 3, S. obliquus and C. vulgaris showed a good ammonia recovery capacity since half of the ammonia concentration was removed by microalgae. Nitrogen removal was directly related to microalgae metabolism activity.
Energies 2020, 13, x FOR PEER REVIEW 7 of 12 swine manure digestate diluted with tap water observed a biomass production of 0.55 g L −1 and NH3 and PO4 3− total removal. Bjornsson et al. [48] suggested that some fresh water micro-and macronutrients could be absent in artificial medium or digestate and this did not permit microalgae growth. For example, Mg 2+ is essential for microalgae growth. Mg 2+ ion is an essential constituent of chlorophyll and its concentration in tap water could be high enough to facilitate microalgae proliferation. Growth rate (µ) values (Table 3) showed that the growth rate of C. vulgaris was faster than the growth rate of S. obliquus in both DIG+ISO and DIG+TW. Values obtained for C. vulgaris were comparable with the ISO control condition. In addition, C. vulgaris showed a longer exponential growth phase compared to S. obliquus (4 days and 2 days, respectively). As reported by Massa et al. [49] who had tested S. obliquus on OFMSW digestate diluted 1:3 in synthetic medium, this strain showed no growth in the presence of a low digestate dilution as obtained in this study. Uggetti et al. [20] showed a S. obliquus growth rate between 0.04 d −1 and 0.90 d −1 on OFMSW digestate diluted 1:5 and 1:10, respectively. Veronesiv et al. [50] tested Chlorella sp. on agro-zootechnical ultrafiltered digestate 1:10 diluted and they obtained a growth rate of 0.2 ± 0.01 d −1 . Growth rates obtained in this study were comparable to those reported by these publications where digestate was autoclaved and filtered before use and microalgae cultivation was performed with air bubbling application.

Nutrient Removal and Chlorophyll a, b and Carotenoid Accumulation
Petri plates test results encouraged batch testing with a supernatant diluted 1:5, with 199 ± 9 mg L −1 initial ammonia concentration in both test conditions: DIG+ISO and DIG+TW. As reported in Figure 3, S. obliquus and C. vulgaris showed a good ammonia recovery capacity since half of the ammonia concentration was removed by microalgae. Nitrogen removal was directly related to microalgae metabolism activity.  The nitrogen removal values obtained in this study were lower than those reported by Massa et al. [49], Ledda et al. [51], Kim et al. [52], Xia and Murphy [19], Cai et al. [21], Cho et al. [22] and Gentili [53] where S. obliquus and C. vulgaris growth was carried out in several kinds of digestate, including OFMSW digestate. In those studies, they reported complete nitrogen removal. However, their microalgae cultivation was performed with air insufflation, so nitrogen removal was partially associated to ammonia stripping with a consequent ammonia loss to the atmosphere. In the present study, without air insufflation, C. vulgaris showed a better performance in terms of ammonia removal compared to S. obliquus in DIG+ISO condition, but the results obtained for ammonia removal on DIG+ISO and DIG+TW conditions were quite similar. Ledda et al. [51] and Nuñez et al. [54] highlighted that only 25-33% of nitrogen lost from medium was recycled in the microalgae biomass, and this percentage was comparable with data obtained with no air insufflation. Nitrate removal showed a similar trend to the Park et al. [55] study, where the nitrate decrease was directly linked with microalgae growth. Phosphate removal showed lower values than those reported in Gentili [53] and Riuz-Martinez et al. [56] where they obtained phosphate removal of 91-99%. Sulfate removal was highest in C. vulgaris; this SO 4 2− ion removal could be associated with CO 2 fixations, O 2 bioproduction, nitrate ion reduction and N 2 fixation; all processes strictly associated with aerobic metabolism. S. obliquus's low sulphate removal efficiency could be associated with no cellular growth and intracellular substance release from biomass during the death stage of the growth curve [57,58]. So, C. vulgaris showed the best phytoremediation effect on digestate compared to S. obliquus during the batch test. S. obliquus chlorophyll a, b (Ch a and Ch b, respectively) and carotenoids analysis ( Figure 4) showed similar values for ISO, TW, DIG+ISO and DIG+TW. On the other hand, C. vulgaris growth on digestate showed a carotenoids increase, while TW favored Ch a and Ch b accumulation. Chlorophyll a and total carotenoids ratio (Ch a/Carot tot ) was measured to identify the stress growth condition linked to carotenoid storage: C. vulgaris' Ch a/Carot tot ratio was 6.6 ± 0.8 and 9.0 ± 0.2 on DIG+ISO and DIG+TW, respectively. These ratios were much higher than the 1.87 ratio reported by Zuliani et al. [9] for C. vulgaris. Chlorophyll and carotenoids analysis confirmed no stress or inhibition effects on C. vulgaris growth on OFMSW-WAS digestate inasmuch as it was in stress growth conditions (i.e., low light penetration or nutrient depletion) inducing Ch a degradation and carotenoids storage, with a strong decrease of Ch a/Carot tot ratio [4,9].  The nitrogen removal values obtained in this study were lower than those reported by Massa et al. [49], Ledda et al. [51], Kim et al. [52], Xia and Murphy [19], Cai et al. [21], Cho et al. [22] and Gentili [53] where S. obliquus and C. vulgaris growth was carried out in several kinds of digestate, including OFMSW digestate. In those studies, they reported complete nitrogen removal. However, their microalgae cultivation was performed with air insufflation, so nitrogen removal was partially associated to ammonia stripping with a consequent ammonia loss to the atmosphere. In the present study, without air insufflation, C. vulgaris showed a better performance in terms of ammonia removal compared to S. obliquus in DIG+ISO condition, but the results obtained for ammonia removal on DIG+ISO and DIG+TW conditions were quite similar. Ledda et al. [51] and Nuñez et al. [54] highlighted that only 25-33% of nitrogen lost from medium was recycled in the microalgae biomass, and this percentage was comparable with data obtained with no air insufflation. Nitrate removal showed a similar trend to the Park et al. [55] study, where the nitrate decrease was directly linked with microalgae growth. Phosphate removal showed lower values than those reported in Gentili [53] and Riuz-Martinez et al. [56] where they obtained phosphate removal of 91-99%. Sulfate removal was highest in C. vulgaris; this SO4 2− ion removal could be associated with CO2 fixations, O2 bioproduction, nitrate ion reduction and N2 fixation; all processes strictly associated with aerobic metabolism. S. obliquus's low sulphate removal efficiency could be associated with no cellular growth and intracellular substance release from biomass during the death stage of the growth curve [57,58]. So, C. vulgaris showed the best phytoremediation effect on digestate compared to S. obliquus during the batch test.
S. obliquus chlorophyll a, b (Ch a and Ch b, respectively) and carotenoids analysis ( Figure 4) showed similar values for ISO, TW, DIG+ISO and DIG+TW. On the other hand, C. vulgaris growth on digestate showed a carotenoids increase, while TW favored Ch a and Ch b accumulation. Chlorophyll a and total carotenoids ratio (Ch a/Carottot) was measured to identify the stress growth condition linked to carotenoid storage: C. vulgaris' Ch a/Carottot ratio was 6.6 ± 0.8 and 9.0 ± 0.2 on DIG+ISO and DIG+TW, respectively. These ratios were much higher than the 1.87 ratio reported by Zuliani et al. [9] for C. vulgaris. Chlorophyll and carotenoids analysis confirmed no stress or inhibition effects on C. vulgaris growth on OFMSW-WAS digestate inasmuch as it was in stress growth conditions (i.e., low light penetration or nutrient depletion) inducing Ch a degradation and carotenoids storage, with a strong decrease of Ch a/Carottot ratio [4,9].