The two Synechococcus strains were distinct in that UHCC0524 reached stationary phase on the third day, while UHCC0527 did not display any growth during the entire experiment. micro-algal strains belonging to the genera Selenastrum, Scenedesmus or Tetradesmus have been extensively used in municipal and industrial wastewater research. Compared to other waste streams such as municipal or poultry wastewater, the hydroponic effluent presents a significantly lower COD and BOD content but higher amounts of nitrogen and phosphorous. Lower COD and BOD content generally allows better light penetration which, combined with rich nutrient concentrations, demonstrates the suitability of hydroponic effluent as a growth medium. The genus Apatococcus represents a highly versatile group of mixotrophic photobionts often establishing symbiotic relationships with lichens and having the ability to survive in terrestrial habitats as free-living cells or in biofilm colonies. To the best of our knowledge, this was the first time an Apatococcus strain was applied for agricultural wastewater remediation, although the presence of this genus in wastewater streams has been previously reported. The Chlorococcum genus has received significant attention due to its growth performance in wastewater streams and ability to synthesize high lipid content potentially suitable for bio-diesel.Likewise, the genus Monoraphidium has been known for the ability to proliferate in industrial waste waters and accumulate saturated fatty acids appropriate for bio-diesel.

The Microcystis genus is commonly identified as a dominant species in harmful algal blooms; however, it is worth noting that not all strains produce microcystins, and in this study a non-toxic strain was used. Its moderate growth in hydroponic effluent suggests that this isolate may not be as robust as previously reported strains. The Nostoc genus represents a group of N2-fixing heterocystous filamentous cyanobacteria that have demonstrated ability to grow in sewage and synthetic waste waters. The unicellular Synechococcus strains are used in wastewater remediation either as immobilized or free-living organisms. In this study, the inoculum of both Synechococcus strains were maintained as free-living cells in low-light intensity, grow table hydroponic and therefore light stress might have inhibited the growth of the UHCC0527 strain. The results of the screening experiment confirmed that the hydroponic effluent is a promising alternative source of nutrients for photosynthetic microorganisms. Several micro-algal and cyanobacterial strains can effectively proliferate in the effluent without any adjustment . The results also showed the potential of the NordAqua Nordic Culture Collection by uncovering several isolates suitable for bioremediation of hydroponic effluents. Considering our extensive knowledge of the performance of UHCC0027 in synthetic and municipal waste waters and previous scale up experience we continued to work with this Finnish isolate of the Scenedesmaceae family in the following up scaling experiments. Besides, the micro-nutrient concentrations of the hydroponic effluent  were within growth promoting ranges for the Scenedesmus genus .

The application of waste streams for the large scale cultivation of micro-algae presents different challenges regarding the elimination of competitors, nutrient composition, and the content of organic matter. Employing suitable pre-treatment methodologies to the source and characteristics of any given effluent can aid the efficient proliferation of the seeded culture. To this end, three different pre-treatments were compared for their potential impacts on the growth of UHCC0027: Coarse filtration , micro-filtration , and bleach . The selection criteria for the applied methods were the presence of native microorganisms in the hydroponic effluent and the feasibility of integration into a pilot scale PBR. A control group was grown in autoclaved BG11 medium. Non-seeded flasks containing only pre-treated effluent were also included to evaluate the impacts of each method on the indigenous microbial community. The results demonstrated the ability of UHCC0027 to proliferate without significant differences between the tested pre-treatments . Additionally, there were no significant differences in the nutrient removal efficiencies with UHCC0027, demonstrating an ability to remove 36–52% and 99% for N-NO3− and P-PO43− , respectively . Interestingly, the non-seeded CF-treated medium demonstrated an increase in OD750 after 7 days of incubation . At day 12, the CF non-seeded medium displayed a significant increase in turbidity  followed by an increase of the total chlorophyll content, suggesting the presence of an indigenous photosynthetic microbial community . Assuming that indigenous microbes were also present in the CF-treated flasks seeded with UHCC0027 but its growth did not significantly differ from the control , it can be concluded that the chosen microalga was able to outcompete the indigenous microbial community.

A negligible increase in OD750 in non-seeded MF and BST flasks by the end of the experiment indicates that these methods are effective in reducing indigenous microbial activity. The effluent used for pre-treatment experiments had an average concentration of 186.17 ± 2.94 mg L− 1 for N-NO3− and 11.77 ± 0.02 mg L− 1 for P-PO43− . The N:P ratio of 35 indicates an unbalanced nutrient ratio that can negatively affect the growth of Scenedesmus species with implications in lipid and protein content . At 22 ◦C, over 90% of the P-PO43− was consumed in the beginning of our experiment, similar to what has been observed previously, although cellular adsorption has been reported to partially contribute to this process. This may have enabled the growth of UHCC0027 for 2 weeks with only residual amounts of P-PO43− in the medium. Thus, uptake of N-NO3− continued until all P-PO43− was consumed. The constant supply of 3% CO2 also favored nutrient uptake by improving the C:N:P ratios during the experimental period until the nutrients were depleted from the medium. At the end of the experiment, microscopic observations of the CF blank samples showed the complex food webs of the indigenous microbial community which were characterized by the presence of several taxonomic entities. In contrast, the microscopic observations of the seeded CF samples demonstrated the dominance of the microalga UHCC0027 . Therefore, since the different pretreatments did not significantly affect the growth or the nutrient removal efficiencies of UHCC0027, and, the culture was able to overcome the presence of indigenous microorganisms in lab scale experiments, a decision was made to combine the BST method with the two step filtration system  for the PBR inoculation. However, the BST method was not applied during the turbidostat mode to ensure the feasibility of daily operation and to facilitate handling large volumes of effluent.

Nonetheless, when compared with the previous setup, the outcome of Setup 2 demonstrated no significant impact in terms of nutrient uptake when the DW of the culture was reduced to one half. Additionally, this result shows that Setup 1 was operated without limitation, presenting an opportunity for increased DW in future experiments. This could improve the removal efficiencies of the system. The following setups, 3 and 4, were planned to evaluate the performance of the culture under varying pH. The transition period between Setups 2 and 3 was the longest undertaken, due to changes in both biomass concentration and pH . A two-step approach was used, whereby the culture was grown to a DW of 1 g L− 1 before the pH was changed. In Setup 3, the pH was lowered to 6.8 and the hydroponic effluent had the highest N:P ratio, oscillating between 32 and 37 . This variability did not compromise N uptake which remained stable at 86.68 ± 6.5 mg g− 1 , but led to a significant reduction of P uptake to 11.84 ± 0.42 mg g− 1 . The removal efficiencies were 31.5 ± 0.9% for N-NO3− and 68.9 ± 5.9% for P-PO43− .By this time, the P-PO43− in the hydroponic effluent was at its lowest concentration which explains the increase in the removal efficiency . The stress induced by different setups, particularly from this transition period onwards, was observed as a gradual bio-fouling effect. In Setup 4, pH was increased from 6.8 to 9 and the N:P ratio varied between 26 and 27. These factors,grow table in combination with several days of high solar radiation  led to a significant increase of P uptake to its maximum value of 21.30 ± 0.96 mg per 1 g DW . During this period, the system met the requirements of the EU directive  for wastewater maximum nutrient discharge for P-PO43− . However, the growth rate was lower than in the previous setups, which contributed to a longer HRT . The contribution of bio-fouling to this outcome should also be considered, since it continued to accumulate throughout Setup 4. Shifting the pH from 6.8 to 9 led to changes in dissolved inorganic carbon species from CO2 to HCO3. The supply of CO2 on demand through set point injection may have created a scenario where different inorganic carbon species were present simultaneously. This detail may have caused extra stress to the culture if the cellular metabolism was activated to favor the absorption of one inorganic carbon source to the detriment of the other.

Nonetheless, loss of growth rate due to a pH as high as 9 has previously been reported for Scenedesmus strains. The N:P ratio of the hydroponic effluent varied significantly between each setup of the PBR . During continuous cultivation, the N:P ratio of the hydroponic effluent varied between 16 and 37, which did not affect the biomass productivity to the extent observed for the chlorophyte C. vulgaris grown in municipal wastewater. Our data suggest no strong correlation between the N:P ratio and the volumetric productivity throughout the course of the experiment. This outcome may result from the high concentrations and constant availability of N-NO3− and P-PO43− in the media as well as the cellular physiological saturation of nutrient uptake. The scenario of nutrient repletion throughout the course of continuous operation is a likely factor in the moderate removal efficiencies observed for N-NO3− and P-PO43− . This, combined with cultivation in a PBR and the steady abiotic conditions provided by the greenhouse infrastructure may also explain the higher biomass productivity observed compared to other studies on Scenedesmus species. Additionally, due to the nutrient replete conditions, it can be hypothesized that the rate of assimilation did not follow a linear relationship with the rate of supply. This trend is described by the theoretical kinetics model of Michaelis-Menten and has been observed by n batch experiments. Despite this, luxury uptake of nutrients may have occurred during the last setup of the PBR run, since the growth rate was at its lowest, but the nutrient uptake was not substantially suppressed . In a continuous cultivation system, the HRT is negatively correlated with the growth rate of the species. If the growth rate decreases, the inflow of media is reduced, which increases the retention time. This trend was verified throughout the course of the experiment, indicating that the transition periods were crucial to stabilize the culture between each setup . Longer retention times yielded higher amounts of biomass to a total maximum of 90.2 ± 0.02 g of DW obtained at the 5th day of the batch operation. During the turbidostat operation the maximum value of total biomass was 67.32 ± 1.85 g of DW, observed in the last period of the PBR run . In our system, the coefficient of correlation between the turbidity sensor and the DW was obtained with a suspended algal culture. Therefore, the gradual accumulation of biofilm presented a threat to the sensor light path and turbidostat operation. For that reason, the PBR was stopped after 36 days of continuous operation and over 1 m3 of hydroponic effluent circulated.