CULTIVATION OF Chlorella vulgaris IN MEDIUM SUPPLEMENTED WITH DESALINATION CONCENTRATE GROWN IN A PILOT-SCALE OPEN RACEWAY

In this study, we investigated the outdoor production of a microalga C. vulgaris cultured in two different media under autotrophic cultivation: Bold Basal Medium (BBM) as the control and BBM supplemented with desalination concentrate (BBM + DC) using open raceway ponds (8 m2). Data were collected on the growth, biomass productivity and biochemical composition. The culture developed in BBM + DC showed a biomass productivity of 6.8 g m-2 day-1, while for the BBM control it was 8.5 g m-2 day-1. Intracellular protein was the main algal component (~28.6%), followed by carbohydrate + fiber (~26.0%) and lipids (~5.0%). The predominant fatty acids were mainly α-linolenic (~19.5%), palmitic (~16.5%) and linoleic (~10.0%) acids. This study demonstrates the feasibility of culturing C. vulgaris in an alternative medium based on DC in order to valorize the desalination wastewater through its application to algal mass production.


INTRODUCTION
The northeast region of Brazil, which has a semiarid climate and frequent droughts, is a vast region (982,563 square kilometers) with brackish groundwater.To face the scarcity of good quality water in some parts of the semiarid region of Brazil, brackish groundwater has been desalinized to make it appropriate for use (Ministério do Meio Ambiente, 2004).Desalination through reverse osmosis is the most attractive solution for water supply (Sánchez et al., 2015).Equipment for the desalination of water based on reverse osmosis has been installed in the community of Uruçu, located in São João do Cariri -Paraíba state.In general, the piped groundwater has a flow intake of around 2-3 m 3 /h, operating with an average recovery rate of 90%.As a result, a waste stream of brine with a high concentration of sodium, calcium and chlorides is produced.
Extensive research studies have been conducted using many types of wastewater for microalgae cultivation, particularly in raceway ponds.As examples, municipal wastewater has been used to grow many types of algae (Pittman et al., 2011), concentrate from anaerobic digestion for the cultivation of Scenedesmus sp.(Tran et al., 2014;Morales-Amaral et al., 2015), effluent from anaerobic digestion to grow Spirulina sp.LEB-18 (Borges et al., 2013), pig wastewater for the cultivation of Spirulina (Arthrospira) (Olguín et al., 2003) and dairy wastewater to grow Chlorella zofingiensis (Huo et al., 2012).Our focus, however, is to produce algal biomass in a medium combined with desalination concentrate (DC), particularly aimed at developing a system which could be integrated with an inland desalination plant.The current model was designed in the community of Uruçu (Paraíba state, Brazil) as a possible site for algal production due to its favorable climate conditions (low seasonal variation and ample solar irradiance) and geography (local source of brackish groundwater and flat topography).
In an open raceway ponds system, two key factors determine the suitability of a microalga for biomass production: the areal productivity (the amount of biomass per unit area (m 2 ) per unit time (day)) (Lawton et al., 2015) and the biochemical composition (Batista et al., 2013;Tibbetts et al., 2015).The microalgae commonly cultivated in raceway ponds include Nannochloropsis sp., Chlorella sp., Tetraselmis sp., Arthrospira (Spirulina) platensis, Dunaliella salina, Scenedesmus sp. and Haematococcus pluvialis (Kumar et al., 2015;Varshney et al., 2015).Most biotechnological research on microalgae has been carried out using Chlorophyta (Varshney et al., 2015).The microalga C. vulgaris (Trebouxiophyceae) used in this study is a unicellular alga, 5 to 8 µm in diameter (Graham et al., 2008).In addition, the genus Chlorella is widely cultured in open raceway ponds under different wastewater conditions (Pittman et al., 2011;Chu et al., 2015;Lu et al., 2015).Furthermore, the Chlorella freshwater microalgae have been targeted for biomass applications due to their high productivity, favorable biochemical composition, cosmopolitan distribution and competitive dominance over other algal species in open culture systems (Wu et al., 2007;Huo et al., 2012).C. vulgaris is also a potential source of lipids, which represent a promising pathway to obtain biofuels from microalgae (Halim et al., 2012;El-Sheekh et al., 2013;Skorupskaite et al., 2015).
The present study was carried out following a project designed to evaluate the performance of C. vulgaris produced in a medium containing DC. Herein, we describe the outdoor batch growth of C. vulgaris in open raceway ponds.The main aim of this research was to investigate the biomass productivity and biochemical composition (e.g., fiber, protein, lipids and fatty acids) of C. vulgaris cultivated in two different media: (1) Bold Basal Medium (BBM) as the control; and (2) BBM supplemented with desalination concentrate (BBM + DC) under the natural climatic conditions of the semiarid region of Brazil.As far as the authors are aware, this is the first description of combining desalination wastewater and algal cultivation in the Northeast region.The importance of this study lies not only in the scientific field, but also relates to the practical application of such a system, not only in the community of Uruçu (Paraíba state) but in the rest of the semiarid regions of Brazil.

Microalgal Species and Inoculum Preparation
The freshwater microalga C. vulgaris was obtained from the Laboratory of Food Biotechnology at the Federal University of Santa Catarina.The alga was maintained in autoclaved BBM, which is suitable for freshwater algae (Nichols, 1973)

Desalination Concentrate
Desalination concentrate was collected from an inland desalination plant, located in the community of Uruçu, São João do Cariri, Paraíba State, Brazil.The brackish groundwater was pumped through a small tubular (100-mm diameter) well (27-m depth) and fed to the desalination tank.The DC samples were obtained after passing the brackish water through three parallel reverse osmosis membranes (FILMTEC TM TW30 4040, Midland, USA) using desalination equipment.The chemical compositions of the groundwater (from the well) and DC (Table 1) were determined according to standard methods described by APHA (2005).

Ponds, Culture Conditions and Growth Measurements
Two above-ground concrete open raceway ponds (area of 8 m 2 , 5.0-m length, 1.6-m width, 0.5-m depth and total capacity of 4,000 L/4 m 3 for each pond) were employed in this study.The outdoor experiments were conducted in 2012 between mid-January and February (austral summer) at an experimental algal farm facility located in the community of Uruçu, São João do Cariri, Paraíba Sate, Brazil (07°23'27" S, 36°31'58"O) with an altitude of approximately 458 m.According to the Koppen classification, the climate in the study region can be described as semiarid (BSh classification) with temperatures varying from 20°C to 35°C during the year.
In order to proceed with the experimental studies in the open raceway ponds, two different media were prepared: (1) BBM (control treatment); and (2) BBM supplemented with 25% of desalination concentrate (BBM + DC) (Figure 1).The ponds were inoculated with exponential-phase inoculum of C. vulgaris grown outdoors in fiber photobioreactors (120 L) placed under natural light illumination and temperature, with saturated air-CO 2 (concentration of 0.03% CO 2 by volume with a flow rate of 1.5 L min -1 ) constantly injected into the photobioreactors.
Approximately 1500 L of fresh culture medium (BBM) and (BBM + DC) was supplied to the open raceway ponds.After 8 days of cultivation, 2000 L of fresh culture medium (BBM) and (BBM + DC) was further added to the two ponds, promoting the microalgal growth until the final working volume (3500 L) was reached.The open raceway ponds were operated in a batch mode (14 days) of cultivation, mixed using a paddle wheel programmed for 10min periods (i.e., turned on/off at regular intervals) 24 h a day, under the exact same conditions (nutrient replenishment, pH, solar incidence, paddle wheel velocity, etc.).There was no external diffusion of carbon dioxide through the mechanical systems during the experimental cultivation.Evaporation losses were at the rate of 1 to 2 cm/day and were replenished with water every morning.Fresh water was used to replace evaporative losses.When the microalgal culture reached a stationary growth phase after 14 days, all of the culture volume from the open raceway pond was pumped into a 4000-L open-top conical tank.Based on previous laboratory results, as reported by Morioka et al. (2014) for Chlorella flocculation, the pH of the culture was elevated to 10.5-12.0 by adding (NaOH) + 0.5-1.0g L -1 of flocculant CaCl 2 .The suspension was manually mixed with a tool (shovel) for 10-15 min and then left to stand overnight (12 h).After the partial sedimentation, the biomass slurry was harvested by continuous centrifugation (Motortronics centrifuge, USA) at 3600 rpm for approximately 3 h.The microalgal pellet was transferred to a dish and dried at 60°C.The dried microalgal biomass samples were transported to the Laboratory of Food Biotechnology (BIOTEC) at the Federal University of Santa Catarina (UFSC) for further analysis to determine the biochemical composition.
Water temperature, pH, dissolved oxygen (DO) and conductivity were measured on-line and recorded continuously every day, using an HI 9828 Series HANNA Multiparameter probe (HANNA, Romania).The Illuminance was obtained using a digital lux meter LD-500 series (ICEL, Brazil).
Growth parameters were measured every day (for cell density and biomass productivity), using the methods described by Lourenço (2006).Cell counting was carried out using a hemocytometer and a compound microscope (Olympus, Germany).For the biomass production, a sample of the microalgae in the growth medium (10 mL suspension) was filtered through glass microfiber GF/C filter paper (Whatmann, New York, USA) and washed twice with distilled water.The paper with attached cells was dried at 105 °C for 1 h, and then kept over a desiccant in a vacuum desiccator overnight.
The chlorophyll a concentration was estimated as described by Ritchie (2006) as an indicator of algal biomass growth.A volume of 25 mL of each sample was filtered through the 0.40 µm glass microfiber GF/C filter paper (Whatman, USA).Chlorophyll a was extracted from filters with 90% acetone (Vetec ® ).Absorbance of the extract was measured at 660 nm (chlorophyll a) and 750 nm (turbidity) with a Hach spectrophotometer (Loveland, CO, USA).Chlorophyll a concentration (µg mL -1 ) was calculated from Eq. 1.
using a nitrogen-to-protein conversion factor of N x 4.78 (Lourenço et al., 2004).
Lipid content -Intracellular lipids were extracted by the Soxhlet method with petroleum ether for 6 h, after acid digestion with 4 N HCl, followed by concentration in a rotary evaporator.The samples were then dried in an oven and weighed (AOAC, 2005).
Fatty acid analysis -The fatty acids composition was determined after converting the fatty acids to their corresponding fatty acid methyl esters (FAME) (IAL, 2005), by gas chromatography using a GC-2014 (Shimadzu, Kyoto, Japan), equipped with a split-injection port, flame-ionization detector and 105 m-long Restek capillary column (ID = 0.25 mm) coated with 0.25 μm of 10% cyanopropylphenyl and 90% biscyanopropylsiloxane.The injector and detector temperatures were both set at 260°C.The oven temperature was initially set at 140°C for 5 min, then increased at 2.5°C min -1 to 260ºC which was held for 30 min.The injection volume was 1 µL, and the split ratio was 10:1.Nitrogen was used as the carrier gas (flow rate 2.2 mL min -1 ) at a constant pressure of 130.3 kPa.Fatty acid methyl esters were identified by comparison with the retention time of individual standards (Sigma, St. Louis, USA).The proportions of the individual acids were calculated from the ratio of their peak area to the total area of all observed acids and expressed as a mass percentage.

Preliminary Studies of the Optimal DC Concentration for C. vulgaris Cultivation
Thirteen microelements were detected in the groundwater and DC (Table 1).The DC is rich in Cl -, Na + and Ca 2+ .In addition, other nutrients (N and P) and trace elements necessary for microalgae growth, including K + , Mg 2+ and Fe 3+ , were detected in the DC.To assess the applicability of DC as a culture medium for C. vulgaris cultivation, the DC was mixed with BBM in different percentages (25%, 35%, 45% and 55% DC).Our previous experimental results obtained in the laboratory (Matos et al., 2015) showed that C. vulgaris was able to grow in media with all of the DC percentages studied; however, the biomass concentration and biochemical composition were significantly affected by the cultivation conditions.For example, at high DC concentrations (i.e., 45-55% DC), the biomass concentration was ~ 130 mg L -1 and the protein and lipid contents were in the ranges of ~ 20.0 and ~ 3.0%, respectively, while upon growing the ( ) where C is the chlorophyll a concentration (µg mL -1 ), V is the volume of solvent (mL), V s is the volume of sample (L), and L is the light path (cm).A is the absorbance and 11.3 is the specific extinction coefficient for acetone.
Biomass productivity calculation (g m -2 day -1 ) was determined according to Bhowmick et al. (2014) Eq. (2): ( ) where AP is the aerial productivity (for carpet area occupied), FCD is the final culture density, ICD is the initial culture density, WV is the working volume, BRT is the batch run time, and CAO is the carpet area occupied.

Analytical Procedures
Moisture -Moisture was determined by drying the sample in an oven at 105°C for 3-4 h (until constant weight) (AOAC, 2005).
Ash content -Total ash content was determined by heating the samples to 550ºC and holding this temperature for 5 h in a carbolite muffle furnace (IAL, 2005).
Fiber content -Total dietary fiber (TDF) content was determined with a total dietary fiber analysis kit (Megazyme International Ireland Ltd, Wicklow, Ireland), which includes enzymatic hydrolysis with α-amylase, protease and amyloglucosidade and is approved by the AACC (Method 32-05-01) and the AOAC (Official Method 985.29).Duplicate samples (approximately 1 g) were suspended in 50 mL phosphate buffer and submitted to enzymatic hydrolysis by incubating with 50 μL of α-amylase at 60°C for 30 min.The pH was adjusted to 7.5, 100 μL of protease was added and the samples were incubated at 60°C for 30 min.In the next step, the pH was adjusted to 4.5, 200 μL of amyloglucosidade was added and the samples were incubated at 60°C for 30 min.Lastly, the fiber was precipitated with 95% ethanol at 60°C, filtered through fritted glass crucibles with a Celite filter and the residue in the crucible was dried in an oven at 105°C, cooled in a desiccator and weighed.
The most notable differences between BBM and DC are related to the NaCl content and electrical conductivity (EC).DC has a NaCl content of around 2412 mg L -1 and EC of 5.6 mS cm -1 , while the corresponding values for BBM are 25 mg L -1 and 1.1 mS cm -1 , respectively.Regarding the concentrations of the major nutrients that are required for microalgae growth (N and P) there are also differences.BBM contains N in the form of nitrate, whereas in the DC, besides nitrate, N is also present as NH 4 + .Concerning the total P, BBM contains 89 mg PO 4 3-L -1 while the DC content is 0.80 mg PO 4 3-L -1 .In addition, BBM presents a N/P ratio of 2:1 and for BBM + DC this ratio is 5:1.Park et al. (2011) have reported that the N/P ratio in wastewater can range from 4:1 to almost 40:1.Overall, the main difference between BBM and DC is that the latter contains high levels of mineral salts.In addition, a small quantity of ammonia (1.35 mg L -1 ) was detected in the DC along with sulfate (102.0 mg L -1 ), compounds which can be toxic to aquatic life (Greenlee et al., 2009).For this reason, C. vulgaris has a limit of tolerance, which is a DC concentration of around 25-30% in the culture medium.It is also important to note that chemical pre-treatment and cleaning is a necessity in most desalination plants, which typically includes treatment with chemicals to avoid biofouling, scaling, foaming and corrosion in desalination plants, which may have adverse effects on aquatic organisms (Lattemann and Hopner, 2008;Melián-Martel et al., 2011).

Water Quality, Algal Growth and Biomass Productivity
Over the course of the outdoor experiment (14 days of cultivation), the water quality was recorded continuously on-line every day (Table 2).During the experiments, the average water temperature was around 33.1ºC at mid-day and ~ 23.0°C at night.Maximum dissolved oxygen (DO) values during the cultivation period were between 10.1 and 12.5 mg mL -1 , with no significant difference between the two pond cultures.The initial and final pH values of the cultures were 8.10 and 9.40, respectively.Illuminance ranged from 1,000 lux (during sunrise and sunset) to 80,000 lux (at mid-day).The values obtained for the variables electrical conductivity, alkalinity, total suspended solids (TSS) and turbidity dynamics for the two ponds were statistically different.For the outdoor pond containing BBM + DC, higher electrical conductivity (2.2 mS cm -1 ) and alkalinity (538 mg CaCO 3 L -1 ) were observed compared with the BBM control (1.1 mS cm -1 and 200 mg CaCO 3 L -1 , respectively).In addition, TSS and turbidity were also higher in the BBM + DC culture.There were no significant differences in the chlorophyll a concentrations for the ponds, with an average of around 350 mg L -1 .Concerning the ammonium composition, the NH 4 -N content for BBM + DC (1.10 mg L -1 ) was 4.5 times higher than that for the BBM control (0.25 mg L -1 ) (Table 2).
Initial results for the growth of C. vulgaris were similar in the two ponds containing the BBM control and BBM + DC (Table 3).Based on the maximum cell density (MCD), C. vulgaris cells were denser in the BBM control (1.2 x 10 6 cel.mL -1 ) than in BBM + DC (8.9 x 10 5 cel.mL -1 ) (Figure 3), while the average specific growth rate was ~0.25 day -1 .The cultures grown in BBM control showed a biomass concentration of around 275 mg L -1 , while for BBM + DC the result was 220 mg L -1 .Actually, the outdoor cultivation of C. vulgaris resulted in an average biomass concentration of 250 mg L -1 and the data reported herein for the biomass concentration of the outdoor culture are comparable with results obtained by Cuello et al. (2015), which ranged from 263 to 403 mg L -1 using Tetraselmis suecica (Chlorophyta) in open raceway ponds.
The values for the biomass productivity of C. vulgaris obtained in the outdoor cultures, i.e., 8.5 g m -2 day -1 using BBM medium and 6.8 g m -2 day -1 using BBM + DC, are close to those previously reported for this type of Chlorophyta species cultured TSS = Total Suspended Solids.Different letters in the same row correspond to significant differences (p<0.05) by Tukey test.
More specifically biomass productivities in raceway ponds across the globe are well illustrated in Table 4.

Analysis of the Biochemical Composition of C. vulgaris
The biochemical compositions of C. vulgaris cultivated in the two proposed media (BBM and  BBM + DC) in the outdoor experiment are shown in Table 5.The dried microalgal biomass presented intermediate moisture content (15.2-18.8%).The ash content observed in the microalgal biomass cultured in BBM + DC (32.8%) was higher than in the BBM control (14.5%), the values being statistically different (P<0.05), which is due to extra precipitated salts, including calcium carbonate, sodium and chlorides that are trapped in the DC-based medium.In addition, these extra salts may interfere in the extraction step for analysis of the proximate composition of the algal biomass (Moheimani et al., 2015).
The organic constituents (e.g., protein, carbohydrate + fiber and lipids) were the major components of the algal biomass studied and differences between the two ponds were observed, with protein being the most abundant component (average 28.6%), followed by carbohydrate + fiber (average 26.0%), and lipid content (average 4.7%).The protein contents of the algal biomass samples were found to be significantly different (P<0.05).The microalgae cultured in the BBM control showed a protein content of 37.2%, almost 2-fold higher than the value for the BBM + DC (20.0%).The reduced protein content in C. vulgaris cultured in BBM + DC is most likely due to a redirection of available energy towards processes such as osmoregulation rather than towards the synthesis of proteins (Lawton et al., 2015;Matos et al., 2017).
With regard to the intracellular lipids, the total lipids varied from 5.8% (BBM control) to 3.7% (BBM + DC) (Table 5).Actually, it is expected that C. vulgaris cultivated in outdoor conditions would show a lipid content of more than 15-20%, while in this case the lipid content was low (<4.0%), and unsustainable for biodiesel production.It has been noted that environmental factors/conditions such as diurnal, climatic and seasonal variations in temperature, humidity and light impart a profound effect on the biomass in outdoor microalgal cultivation (Varshney et al., 2015).In this regard, the harsh environmental conditions imposed in the present outdoor experiment [i.e., high temperatures (~33°C during the day), high solar radiation (~90 klux mid-day) and nutrient medium containing brine (high salt content)], will have affected the growth and biochemical composition of the C. vulgaris.Regarding the salinity, it is expected that BBM and BBM + DC have a NaCl concentration of around 0.4 mM and 10.3 mM NaCl, respectively, that is, the BBM + DC is 25 times more saline than the BBM medium.However, the salinity of BBM + DC can be considered as moderate, being much lower than that of seawater (0.5 M NaCl).Takouridis et al. (2015) demonstrated the feasible growth of the freshwater microalga Chlamydomonas reinhardtii under high salinity conditions (0.3-0.7 M NaCl) via the selective breeding method of genome shuffling.Nevertheless, only a few microalgal strains can support high salinity, for instance, Dunaliella, a green unicellular microalga isolated from high salinity water bodies with NaCl concentrations exceeding 3.0 M NaCl (Borowitzka and Huisman, 1993).
As shown in Table 6, C. vulgaris cells that were cultured in BBM tend to produce higher levels of PUFAs (mainly C18:2 ω6 and C18:3 ω3), an average of 32.2%, than culture developed in BBM + DC.Furthermore, the PUFA/SFA ratio was higher for the BBM control (1.67) than for BBM + DC (1.21).On the other hand, C. vulgaris cultivated in BBM + DC appears to have a tendency to synthesize more SFAs (mainly C16 and C18:0), consistent with previous data on this fatty acids fraction (Matos et al., 2015).Functionally, higher synthesis of SFAs in C. vulgaris can be explained by an adaptive osmoregulatory mechanism to cope with rapid or gradual changes in salinity that are associated with algae-cell-membrane permeability (Takagi and Karseno, 2006;Lawton et al., 2015).
of DC as a substrate for microalgae cultivation using open raceway ponds was investigated.The microalga C. vulgaris was able to grow in DC, but requires a maximum of 25% of DC concentration.Growth of C. vulgaris cells was similar in BBM and BBM + DC with a biomass productivity of an average of 7.6 g m -2 day -1 .Cultures developed in BBM showed a protein content of 37.2% while in BBM + DC the same metabolite was 20.0%, indicating that C. vulgaris cultivated in DC shifts its metabolism to osmoregulation rather than towards the synthesis of proteins.Finally, this study demonstrated the use of available DC as nutrient feedstock, which could reduce the need for fresh water and external nutrients.This can be considered as a step forward for the valorization of brine management through algal mass cultivation.

Figure 1 .
Figure 1.View of the open raceway ponds for microalgae cultivation at an experimental algal farm facility located in the community of Uruçu, São João do Cariri -Paraíba State, Brazil.A -Photograph of the ponds (5.0-m long, 1.6-m width and 0.5-m depth; totaling 8 m 2 and total capacity of 4,000 L/4 m 3 each pond); B -Cultures of C. vulgaris in BBM (control, left) and BBM supplemented with DC (BBM + DC, right).

Figure 2 .
Figure 2. Effect of different concentrations of desalination concentrate (DC) on protein, lipid and biomass concentration of C. vulgaris under laboratory conditions.

Figure 3 .
Figure 3. Growth curve of C. vulgaris cultivated in open raceway ponds with working volume of 3,500 L during 14 days; (Δ) Control treatment (BBM) and (•) BBM supplemented with DC (BBM + DC).
(2012) * Cultures developed in the control treatment (BBM) and ** cultures developed in the BBM supplemented with DC (BBM + DC); * A = area, ** D = depth of the raceway pond, *** V = velocity of the paddle wheel rotation.

Table 1 .
Compositions of groundwater and desalination concentrate (DC) from the community of Uruçu, Paraíba State, Brazil.

Table 2 .
Median ± standard deviation of water quality variables in the open raceway ponds with respectively culture medium -BBM (control) and BBM supplemented with DC (BBM + DC) -measured during 14 days of operation.
MCD = maximum cellular density; Values shown are averages of three ± the range.

Table 3 .
Maximum cell density, the growth rate and biomass concentration of C. vulgaris cultured in open raceway ponds in control treatment (BBM) and BBM supplemented with DC (BBM + DC).

Table 4 .
Comparison of biomass productivities (g m -2 day -1 ) in open raceway ponds between this study and literature values.

Table 5 .
Chemical composition of C. vulgaris cultivated in open raceway ponds.