EFFECTS OF WATER RECIRCULATION ON PILOT - SCALE MICROALGAE CULTIVATION USING FLUE GAS CO 2 By Carly D aiek A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Biosystems Engineering - Master of Science 2020 ABSTRACT EFFECTS OF WATER RECIRCULATION ON PILOT - SCALE MICROALGAE CULTIVATION USING FLUE GAS CO 2 By Carly D aiek This study investigates the effects of media recirculation on microalgal species C. sorokiniana growth in a pilot - scale algae photobioreactor ( APB ). Two culture conditions, freshwater and recirculated boiler water, were conducted on the APB to determine the ef fect of recirculation on algal growth. The results showed that microalgae cultivation under recirculation conditions was stable over a period of four months. Biomass productivities during the 1 st through 4 th months of recirculation (0.26, 0.23, 0.20, and 0 .18 g L - 1 d - 1 , respectively) were not significantly different from the culture on freshwater (0.22 g L - 1 d - 1 ). Furthermore, the relationship between eukaryotic and bacterial domains remained consistent throughout the four months of recirculation (8 0.7 , 8 7 .1 , 83 .1 , and 82 .1 %, respectively and 19 .2 , 12.8 , 1 6.9 and 1 7.8 %, respectively). This was not significantly different from the abundance of each domain in freshwater cultivation (8 3.7 % eukaryotic and 16 .2 % bacterial). A 1 m 3 photobioreactor was then envisioned for mass, energy and exergy analys e s. The mass balance analysis concluded that a 98% reduction in freshwater usage and 25% reduction in nutrients could be achieved during cultivation operating under recirculation condit ions for a year, while maintaining a biomass productivity of 1.2 kg wet algal biomass and 0.4 kg CO 2 sequestered per day. Both systems require an energy input of 219 kWh unit - 1 d - 1 . The exergy balance analysis concluded that without considering solar irrad iation , the rational exergy efficiency of the culture with water recirculation was more than double that of freshwater. iii ACKNOWLEDGMENTS I wish to express my sincerest gratitud e to all who have helped me achieve this thesis. To my advisor, Dr. Yan (Susie) Li u, whose patience, intelligence, and kindness continue to inspire me. I am incredibly fortunate to have started my professional career under her guidance. Dr. Wei Liao, for serving on my committee and providing expertise , enthusiasm and countless opportunities throughout both my undergraduate and graduate degrees. Dr. Karen Draths , for serving on my committee and providing valu able time and insight to my thesis topic. Mr. Nathan Verhanovitz , for his help and support at the MSU T.B. Simo n Power Plant . To my fellow laboratory members , Ashley Cutshaw, Annaliese Marks, Adam Smerigan, Douglas Clements, Henry Frost, Jack Blackhurst, Meicai Xu , Sibel U ludag - Demirer , and Yurui Zheng, for their knowledge, support , humor and friendship . To Tammy Wells - Wilkinson and Jimmy Larson, for encouraging my development in scientific research and for their guidance and friendship. Finally, I would like to thank my family and friends for their love and encouragement during my time at Michigan State Universi ty. iv TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ...................... vi i LIST OF FIGURES ................................ ................................ ................................ ................... vii i CHAPTER 1: LITERATURE REVIEW ................................ ................................ .................... 1 INTRODUCTION ................................ ................................ ................................ ....................... 1 EFFECTS OF MICROBIAL COMMUNITY ................................ ................................ ............. 5 Micr oalgal strain selection and Chlorella sorokiniana ................................ ........................ 6 Effect of bacterial community ................................ ................................ ................................ 8 EFFECT OF CULTURE MEDIA ................................ ................................ ............................. 10 Effect of accumulation and inhibitory compounds ................................ ............................. 10 Effect on growth and composition ................................ ................................ ....................... 14 SUMMARY OF KNOWLEDGE GAPS ................................ ................................ .................. 15 OBJECTIVES AND HYPOTHESIS ................................ ................................ ........................ 16 CHAPTER 2: CONTINUOUS MICROALGAE CULTIVATION ON FLUE GAS IN A WATER RECIRCULATING PHOTO - BIOREACTOR SYSTEM ................................ ...... 17 ABSTRACT ................................ ................................ ................................ .............................. 17 INTRODUCTION ................................ ................................ ................................ ..................... 18 MATERIALS AND METHODS ................................ ................................ .............................. 20 Algal Assemblage ................................ ................................ ................................ .................. 20 Pilot photobioreactor system and operations ................................ ................................ ....... 20 Chemical analysis ................................ ................................ ................................ ................. 21 Microbial analysis ................................ ................................ ................................ ................. 22 DNA Extraction ................................ ................................ ................................ .................. 22 Illumina preparation and sequencing ................................ ................................ ................ 22 Statistical analysis ................................ ................................ ................................ ................. 23 Mass balance analysis ................................ ................................ ................................ .......... 24 Energy balance analysis ................................ ................................ ................................ ....... 24 Exergy analysis ................................ ................................ ................................ ..................... 25 RESULTS AND DISCUS SION ................................ ................................ ............................... 28 Effects of water recirculation on algae growth, nutrient consumption, and biomass composition ................................ ................................ ................................ ........................... 28 Effects of water recirculation on algal assemblage ................................ ............................ 30 Mass and energy balance of the water recirculating photobioreactor system ................... 38 Exergy analysis of the water recircu lating photobioreactor system ................................ ... 41 CONCLUSION ................................ ................................ ................................ ......................... 42 ACKNOWLEDGEMENTS ................................ ................................ ................................ ...... 43 CHAPTER 3. CONCLUSIONS AND FUTURE WORK ................................ ....................... 44 CONCLUSIONS ................................ ................................ ................................ ....................... 44 FUTURE WORK ................................ ................................ ................................ ...................... 45 v APPENDICES ................................ ................................ ................................ ............................. 46 APPENDIX A: TABLES AND FIGURES ................................ ................................ ............... 47 APPENDIX B: R CODE FOR PLOTTING AND ANALYSIS ................................ ............... 67 REFERENCES ................................ ................................ ................................ ............................ 86 vi LIST OF TABLES Table 1 : Biomass composition under recirculation and freshwater conditions .............. .. .... . ....... 30 Table 2: Resources saved of the envisioned 1000 L algae photobioreactor for 1 - year operation under recirculation conditio ns ........... ............. ... ....... ............. ........................... .. ......... 40 Table 3: Energy balance of the envisioned 1000 L algae photobioreactor ........................ ... .. ...... 4 1 Table 4 : C. sorokiniana biomass composition ... . ......... ......................... ............. ... ..... .......... .. ....... 48 Table 5 : C. sorokiniana CHNS composition ..................... .......... .................................... .. .......... . 49 Table 6 : C. sorokiniana fatty acid composition as % of dry matt er ....... . .......................... . ..... ..... 49 Table 7 : C. sorokiniana fatty acid composition as % of total fatty acid s .. .. .... ................... .. ........ 50 Table 8 : Algal supernatant and boiler water composition ... .................................. ............... . .... ... 50 Table 9 : Biomass data under recirculation conditions ... .... ..................... ............................. .. .. ..... 51 Table 10 : Biomass data under freshwater conditions ... ................................. ........................... .. .... 52 Table 11 : Operational data for recirculation cultivation ... ........ . ........... ........ .. ................ .............. 53 Table 12 : Operational data for freshwater process . .. ... ............... ....................... ....... . ................ ... 56 Table 13 : Illumina OTU classification key ............................... .............................. .. ..... ............... 58 Ta b le 14 : Illumina sequencing abundance under recirculation conditions ............ .. .......... . .......... 59 Table 15 : Illumina sequencing abundance under freshwater conditions ... .......... ..... .. .................. 60 Table 16 : Process compounds, mass flow rate, temperature, chemical exergy rate, physical exergy rate, and total exergy rate of individual compounds for microalgae cultivation under recirculation conditions ... .............. ........... ... .......... ... .................... .. ................ 65 Table 17 : Process compounds, mass flow rate, temperature, chemical exergy rate, physical exergy rate, and total exergy rate of individual compounds for microalgae cultivation under freshwater conditions .......... . ...... .... ........... ........... ... ............. ................ .. .................... 66 vii LIST OF FIGURES Figure 1: Biomass productivity and n utrient consumption .................................. .. ....................... 29 Figure 2 : Relative abundance of bacterial communities ............... .. ............... .. .......................... ... 32 Figure 3 : Non - metric multidimensional scaling ( NMDS) analysis of microbial community relative abundance for the APB system .......................................................... .. .............. ............. .. 34 Figure 4: Mass balance of the pilot APB unit ..................................................... .. ...................... .. 39 Figure 5 : Rational exergy efficiencies of freshwater and water recirculation cultivations .... . .. .. . 4 1 Figure 6: 100 L photobioreactor set up. ..................................................... ... ..................... .. ......... 47 Figure 7: C. sorokiniana and microbial community .................................... ....................... .. ........ 47 Figure 8: Overall microbial community for freshwater and recirculation samples ......... .. ........... 61 Figure 9 : M icrobial community abundance ........................ .............. ......................... .. . ... ............ 61 Figure 10 : Heatmaps of microbial community abundance ........................... ......... .. ...... ............... 62 Figure 11 : Rarefaction curves ................................................................... ... ....... .. ........................ 62 Figure 12 : Scatter plots for community indices ................................... ........ .. ........................ ....... 63 Figu re 13 : Boxplots for microbial community indices comparing recirculation and freshwater conditions ............................................................................................... .. ...................................... 64 Figure 14: Dendrogram of individual recirculation (S8 - 22) and freshwater (S23 - 35) samples .............................................................................................. ... ... .. ..................................... 64 1 CHAPTER 1: LITERATURE RE VIEW INTRODUCTION Carbon dioxide ( ) is a major greenhouse gas and principal contributor of global climate change. Emissions of are largely caused by anthropogenic activities such as burning fossil fuels, electricity and heat production, and deforestation. As the global population continues to increase, global energy consumption has also increased accordingly. Worldwide energy demand is driven by a grow ing global economy along with higher heating and cooling requirements in certain areas of the world [1] . Over half of the growth in energy needs in 2018 was due to a higher electricity demand [1] . Rising energy demands is a main reason for the continued increase of levels in the atmosphere. While many methods for mitigation exist, such as implementation of alternative energy sources, demand side management and climate engineering, sequestration represents a nother possible solution to decrease atmospheric and mitigate global warming. sequestration can be defined as the capture and long - term storage of carbon that would otherwise be emitted to or remain in the atmosphere [2] . Mitigation of has been attempted through a number of sequestration methods, utilizing a variety of chemical, physical and biological processes . One such method is biological carbon capture through microalgae cultivation. Microalgae are photosynthetic organisms with the ability to convert atmospheric into glucose for their growth using solar energy. Microalgae have high capabilities to fix and generate biomass, offering attractive advantages over other terrestrial sequestration methods. They are able to duplicate cell biomass 100 times faster than terrestrial plants and are able to fix 10 - 50 times more efficiently [2] . It has been approximated that 1.83 kg of can be fixed for every 1 kg of biomass produced [2] . Microalgae remain photosynthetically efficient even 2 under a range of concentrations [3] . Furthermore, the cultivation of microalgae biomass does not require arable land [3] . In addition to capture, microalgae biomass is characterized by high protein, lipid, and carbohyd rate content, which provides a good source for value - added products including pharmaceuticals, biofuels, and nutritional supplements [3, 4] . As a result, algae cultivation offers an attractive solution for capture from major c arbon - emitting sources such as power plants. Photoautotrophic microalgae cultivation for CO 2 sequestration requires several components for the successful accumulation of biomass. To maintain a high yield, systems require light for photosynthesis, as a carbon source, a constant supply of several inorganic nutrients, and water [2, 5] . The major sources of used for cultivation are from air or flue gas. is available with atmospheric concentrations of 0.03 0.06% (v/v) or with power plant flue gas rang ing from 6 15% (v/v) [2, 6] . Light requirements are provided through natural sunlight or illumination provided by artificial fixtures. Inorganic nutrients, such as n itrogen (N), p hosphorous (P), and i ron (Fe), can be provided directly to the culture or may be obtained through growth in nut rient - rich water sources, such as wastewater. Microalgae cultivations can be performed in a variety of water sources including freshwater, saltwater or brackish water, depending on the strain in question. Large - scale microalgae cultivations are typically performed under controlled conditions in either open or closed algae photobioreactor ( A PB) systems. Open systems, such as raceway and open ponds, can be mixed or unmixed ponds used for the mass cultivation of algae with constant exposure to the environment [7, 8] . These systems are characterized by minimal capital and operating cost due to lower energy inputs , such as low mixing requirements and use of natural sunlight for illuminance [4, 7] . A major disadvantage of open ponds is the large area 3 requirement for scaling up and signifi cant CO 2 losses to the atmosphere [5, 7] . Additionally, as open ponds are continuously exposed to the environment, these systems are more suscepti ble to contamination and adverse weather conditions [9] . Thus, location is an essential factor in the design of open ponds systems. Furthermore, they are more challenging in terms of controlling growth pa rameters including light exposure, temperature, and evaporation [5] . Microalgae cultivations performed in APB systems are typically grown i n closed tubes or bags, which both reduces exposure to the environment and allows for greater control of growth parameters [7, 9] . CO 2 is also utilized more efficiently in these systems [5] . Artificial light sources, such as light emitting diodes (LEDs), are commonly used in APB systems, allowing for increased light intensity and exposure. Limited exposure to the environment decreases the probability of contamination, complications from adverse weather , and evaporation [5] . These characteristics correlate to a smaller area requirement and versatile location options. However, APB s tend to have issues with biofilm accumulation, overheating, and cleaning issues [7] . Most importantly, they incur very high c apital and operating costs [7] . Despi te these disadvantages, APB s are more efficient than open pond systems in terms of controlling growth parameters while also offering higher biomass productivity [5] . Several major technical challenges on large - scale algal cultivation hinder commercial algal production, including low algal biomass yield in outdoor conditions, lack of long - term stability, and high water and nutrient requirements . Additionally, the area required for consumption to balance industrial emissions is extremely large, leading to technical and economic limitations for large - scale microalgae cultivation [10] . Improvements in operational efficiency within APB operations is essential for technical and economic feasibility of this technology on a commercial scale. 4 One of the main operational limitations is the cost and environmental i mpacts related with the consumption of freshwater and nutrients [11] . These resources are key components in algae cultivation, as a constant supply of water and certain inorganic nutr ients are required to achieve and maintain high biomass productivity. However, it is predicted that a significant cost reduction of at least 50% could be achieved if nutrients and water are obtained at a lower cost [11] . Both freshwater and nutrient usage in microalgae cultivation pose challenges to system sustainability. Freshwater is a natural resource that is becoming increasingly scarce, as freshwater aquifers are currently facing unsustainable rates of extraction [12] . Furthermore, life cycle assessments (LCAs) often find that the life cycle burden of microalgae cultivation come s from nutrient production occurring upstream of algae cultivation facilities [12] . Therefore, maximizing water and nutrient use efficiency is a signif icant factor in improving the overall feasibility of the technology. Current methods used to address water and nutrient challenges involve incorporating alternative water sources into microalgae cultivation systems. Potential alternative water sources incl ude saltwater, brackish water, or recycled freshwater [3, 11, 12] . Certain strains of algae are known to tolerate high concentrations of salt and can be used in systems utilizing saltwater [4] . However, large requirements of saltwater may also be an issue [12] . The outcome of drawing from saltwater aquifers, a nearly untouched resource, is unknown and carries a high risk to coastal environments which are known to be highly productive ecosystems [12] . Additionally, saltwater would require commercial systems to be located near coastal regions to reduce the distance required for water transport [12] . The alternative option would require long distance pipelines that may drastically increase cost in addition to possible environmental and social impacts [11, 12] . Another option is to use wastewater to reduce reliance on freshwater while also provid ing nutrients to the algae [4, 5] . It would also provide 5 biological cleaning options for municipal wastewater and would lower environmental impacts and treatment costs [12] . Drawbacks associated with cultivation in wastewater i nclude fluctuating nutrient level concentration, increased turbidity causing light penetration issues, and rigorous use of toxic chemicals [13] . The final practice involves the recirculation of freshwater and culture media. Recirculation can reduce freshwater requirements while also reducing nutrient usage. An LCA study performed by Yang et al. in 2011 found that when harvesting water was completely recycled, nutrient usage decr eased by about 55% and freshwater usage decreased by 84% [12] . In 2020, Fret et al. demonstrated a 77% decrease in water footprint and 68% reduction in nutrients using media recirculation [14] . However, the full pote ntial of media recirculation across many microalgal species has yet to be completely explored in the field of large - scale cultivation. It is important to note that in order to maintain economic viability, large quantities of biomass and value - added product s derived from algal biomass remain an essential component in microalgae cultivation systems. Although recirculation can reduce water and nutrient usage, the impact of recirculated media on microalgae growth and biomass, and thus the correlated economic va lue of biomass, is still widely unknown. To determine whether or not this approach is feasible, research must be completed to evaluate the effect of reused resources on algal growth and composition in pilot and large - scale operations. EFFECTS OF MICROBIAL COMMUNITY In order to fully exploit the industrial potential of algal biomass in CO 2 sequestration, operational efficiency must be achieved through reduction of nutrient and water usage. Although there is much research on APB design and biomass productivit y optimization, minimal research was conducted on the integration of recycled water and nutrient sources into APB s. As the 6 overall system operation and effects on biomass composition is still largely unknown, performance indicators should be investigated t o further determine feasibility. Microbial community structure and function has the potential to indicate system operational status, including systematic stability. Although knowledge on complex microbial relationships between microalgae and bacteria is st ill limited, monitoring the presence and abundance of certain species can reveal potential pathogens that may be responsible for system failures [8, 15, 16] . Additionally, monitoring microb ial community is useful in identify ing helpful taxa that may correlate to improvements in system performance [8, 17, 18] . M icroalgal strain selection and Chlorella sorokiniana M icroalgae are unicellular, photosynthetic organisms that live in a wide range of aquatic environments such as lakes, rivers, ponds, oceans, and are even known to be found in certain types of industrial effluents [4, 5, 7] . Microalgae can be prokar yotic, such as cyanobacteria, or eukaryotic like green algae [3] . In order to sustain growth, microalgae require light energy to convert water and CO 2 into biomass through photosynthesis. In addition to light, water , and CO 2 , microalga l growth also requires both macro and micronutrients. Carbon is the most important element for microalga l nutrition, with dried algal biomass containing approximately 50% carbon, followed by nitrogen and phosphorous, which account for 10 - 20% of algae biomass [11, 19] . Other commonly required elements include macronutrients Na, Mg, Ca and K and micronutrients Mo, Mn, B, Co, Fe, Zn and other trace elements [4] . Although approximately 40,000 different species of microalgae have been reported, only a handful of those st rains are considered feasible for mass cultivation [2, 4] . Mass cultivation of microalgae requires a strain that can tolerate a wide range of conditions such as temperature, pH, 7 salinity and light intensity. Robust strains that are commonly employed in APB s for seques tration include, but are not limited to , Nannochlorophsis sp., Dunaliella sp., Scendesmus sp. and Spirulina sp. [5, 15] . Additionally, Chlorella is a species with industrial potential because it can grow both photoautotrophically and heterotrophically with high biomass concentration [20] . This microalgae is also commercially important, with global annual sales of greater than $38 billion USD [20] . Chlorella is known to produce value - added chemicals, such as ß1,3 - glucan and carotenoids, and shows promise for biofuels production under heterotrophic conditions [20] . Although many species of Chlorella have been studied for large - scale algae c ultivation, one recent strain of interest is the species Chlorella sorokiniana . Originally isolated by Sorkin and Myers in 1953, C. sorokiniana is a type of green microalgae commonly used in large - scale APB s due to its high photosynthetic productivity and ability to grow at temperatures up to 38 - 42 C [21, 22] . C. sorokiniana is small in size (2 - 4.5 µm diameter) and often found in freshwater and soils [23, 24] . It is one of the only known species of Chlorella that to lerates high temperatures and light intensity, making it beneficial for many types of cultures [21, 22, 25] . It has also been shown to grow in wastewater, under conditions that other algal species may find unfavorable [26] . Furthermore, it has shown resistance to high concentrations of and , compounds that are typically found in power plant flue gas emissions and exhibit potential toxicity to some microalga l spe cies [21] . Under photoautotrophic conditions, cell doubling times are found to be as low as 4 - 6 h [24] . On average, C. sorokiniana is composed of 40% prote in, 30 - 38% carbohydrate and 18 - 22% lipid [24] . Prior research has shown that this robust species has industrial potential and is well suited for large - scale production in air - and liquid - mixed photobioreactors, while also producing compounds of commercial interest including antioxidants (i.e., carotenoids ) [24] . 8 Effect of bacterial community Microalgae can either be cultivated as a pure culture, containing only the species of in terest, or as a co - culture, containing microalgae and other micro - organisms simultaneously. P ure cultures are highly impractical because they are difficult to maintain and are characterized by high capital and operating costs, thus, co - cultures are becoming more common in the field of mass microalgae cultivation [16] . When considering a co - culture system, the analysis of both selected microalga l strain and the overall microbial community is of great importance. Although bacteria have often been considered contaminants that have the potential to inhibit or kill microalgae cultivations, algae - bacteria interactions have many possible effects [17, 18, 27] . In nature, many algae - bacteria interactions occur, with relationships ranging from mutualism to parasitism [16, 17] . Many relationships are still unexplored, especially under lesser - known conditions such as recirculated growth medium. In both natural and industrial processes, there is evidence of microalgae and bacteria living toge ther in complex communities. Many of those described in engineered systems are also of the same genera found in natural environments [18] . Although studies are limited regarding microalgae - bacteria relationships in APB co - culture systems, the presence of several taxa have been documented. An analysis of several large - scale system studies showed that Proteobacteria , part icularly Gammaproteobacteria , were associated in all of the microalgae cultivation communities studied [15] . The study also found the presence of several common bacterial orders, Cytophagales, Flavobacteriales, Pseudomonadales, Burkholderiales, Caulobacterales and Rhodobacterales, though these bacteria were not consistent over all studied systems [15] . The performance of microalga l species is highly affected by various factors such as pH, temperature, nutrient concentration, and light intensity [16] . It is thought that the presence of 9 bacteria in co - culture systems may lead to more robust communities that can better withstand environm ental challenges through communication and division of labor [16] . In general, co - cultures have shown improvements in yields of biomass, lipid s, and other value - added products in comparison to pure cultures [16] . This suggests a positive effect of algae - bacteria symbiosis on alga l gr owth. For instance, o ne study showed that of 326 alga l species studied, 171 species required an external supply of vitamin B 12 [17] . It has also been shown that some bacterial species are known to supply vitamin B 12 to algae in exchange for fixed carbon [8] . Similarly, other bacterial groups may help regulate available nutrients like iron, nitrogen, and phosphates or by releasing growth ho rmones [8] . Although limited, there have been several studies regarding positive relationships be tween Chlorella and several bacterial groups. A review performed by Lian et al discussed a number of bacteria that have been found beneficial to C. vulgaris , including members of the genera Bacillus, Flavobacterium, Rhizobium, Hyphomonas and Sphingomonas [15] . The review also discussed the species, B. pumilus ES4, which has been shown to promote C. vulgaris growth by providing fixed atmospheric nitrogen [15] . Amavizca et al found similar results in a C. sorokiniana co - culture , where B. pumilus ES4 and Azospirillum brasilense Cd were shown to remotely induce increases in total lipids, carbohydrates, and chlorophyll a [28] . Another Chlorella species, C. ellipsoidea , showed increased cell density when accompanied by Brevundimonas sp., while C. sorokiniana IAM C - 212 was shown to have an increased growth rate when grown with Microbacterium trichotecenolyticum [15] . However, within their natural environment, microalgae are still at risk of viruses, parasites, and bacterial pathogens, though many have not been identified [8, 15] . Additionally, there is a greater risk of infection and inhibition from bacteria, fungi, and viruses found in greater 10 concentrations in recycled waters [11] . Co - inhabiting species ma y compete for existing nutrients, resulting in decreased growth of the algae [8, 17] . Another potential thre at is through bacterial parasitism of algae, where algae cells are lysed by enzymes , allowing bacteria to use intracellular compounds of algae as nutrients [17] . Lian et al found that rot symptoms are commonly due to gram - negative members belonging to the genera Alteromonas, Cytophaga, Flavoba cterium, Pseudomonas, Saprospira, Vibrio and Pseudoalteromonas [15] . In addition to bacterial pathogens, some types of algae can also be parasitic [8, 17] . An algaelytic protist, Pseudobodo sp. KD51 s, caused more than a 50% decrease in chlorophyll content of C. vulgaris within three days of inoculation [15] . EFFECT OF CULTURE MEDIA As freshwater and nutrients are limited resources, the integration of recycled media into APB systems is a potential solution [11] . However, the recirculation of freshwater may result in the accumulation of numerous compounds at a level not normally exhibited in cultures frequently replenished with freshwater. Many studies have investigated the use of algae in treating various water sources, suggesting microalgae are tolerant to a wide var iety of water sources and compounds. Research has also been conducted on the use of recirculated freshwater on a variety of commonly cultivated microalgal species, but specific knowledge of the effects o f C. sorokiniana on recycled media is limited. As tol erances and thresholds to various compounds differ between microalgae species, it is still widely unknown how cultures of C. sorokiniana respond to recycled media. Effect of accumulation and inhibitory compounds While alga l cells require a certain amount of nutrients and minerals, the overabundance of any one compound may negatively impact growth. For instance, C, N and P are the most 11 important nutrients required for microalgae cultivation, but an oversupply can result in increased stress and reduced growt h [29] . Similarly, micronutrients, such as Fe, also h ave a supply threshold, however, it is much narrower than that of macronutrients [29] . For example, in a study performed by Wan et al , Fe was found to be beneficial to C. sorokiniana at concentrations up to 10 - 5 mol L - 1 , but was toxic at 10 - 3 mol L - 1 [30] . An investigation on NaCl concentration found that C. sorokiniana tolerated levels up to 0.3 M, but also showed a decreased growth rate [31] . Heavy metals such as cadmium (Cd), lead (Pb), and mercury (Hg) are unnecessary for algae growth and have been shown to negatively impact cells at very low concentrations [15, 32] . For example, a study performed by Carfa gna et al found that the algal cell structure and physiological characteristics such as growth, photosynthesis, respiration and enzyme activities were affected in a strain of C. sorokiniana when exposed to certain levels of Pb and Cd [2 3] . Additionally, the study found that Pb and Cd induced a reduction in total chlorophyll content and decreased soluble protein. Similarly, Liang et al found that C. sorokiniana was able to tolerate levels of Pb (total), Copper (Cu) and Cd at levels of 0.249 mg/L, 0.485 mg/L, and 46.108 mg/L, respectively [33] . Both studies found that C. sorokiniana showed a high tolerance to Pb over the other studied metals, which is likely due to the intra - and extracellular mechanisms possessed by microalgae that prevents metal toxicity [23] . Other heavy metals like zinc, which is beneficial to algae growth in small amounts, may inhibit productivity after a certain concentration [34] . Spence observed inhibition in C. sorokini ana under recycled media conditions but was unable to identify if inhibition was due to the accumulation of zinc or inhibitory secondary metabolites [34] . Accumulation is noteworthy for systems using recycled media, since even a minor accumulation of nonessential compounds can be toxic to algae [35] . 12 Auto - inhibitory compounds are another potential challenge for systems implementing recycled growth media. Auto - inhibitory comp ounds are naturally occurring substances released growth of other species [32, 36] . Auto - inhibitors are typically present in ultrahigh density microalgae cultures, typically characterized by concentrations of at least 10 g cell mass L - 1 , though concentrations at this level are rarely reported in photoautotrophic cultivations [37, 38] . Several specie s of microalgae have been shown to release extracellular compounds with inhibitory or algicidal properties. C. pyrenoidosa , for example, was found to produce polyunsaturated fatty acids, linoleic and linolenic acid, which resulted in inhibitory effects on growth [39, 40] . A genus of algae commonly used as aquaculture feed, Nannochloropsis, is known to release a thick and multilayered parent cell wall during cell division, which may potentially reduce culture growth and produ ctivity using recycled media [29] . Cell wall remains c aused the formation of aggregates in the culture, which are thought to entrap cells, bacteria and debris, leading to unsuitable growth conditions [41] . Additionally, the accumulation of dissolved organic matter can be conducive to algae contamination and can also inhibit algae growth at certain thresholds [38] . In general, the manner in which these substances inhibit growth is widely unknown and many of the substances involved have not been fully characterized [32, 40] . As previously discussed, co - culture systems have been proven to have symbiotic effects on the growth of algae and are often easier to maintain than pure cultures, which explains why many systems today employ co - cultured microbia l communities [42] . However, mixed cultures also present the issue of allelopathy, in which co - inhabiting organisms produce biochemicals, known as allelochemicals, capable of influencing the growth and survival of other organisms [43] . Similar to auto - inhibitory compounds, very few algicidal metabolites have been 13 characterized to date [44] . Allelopathy has frequently been studied, specifically for the role of algicidal bacteria in alga l blooms and can provide insight on compounds that may also be present in mixed cultures. An example of this is seen in the relations hip between C. vulgaris and bacteria Pseudomonas. When allowed to grow at high cell concentrations, the bacteria is found to express self - regulation and inhibitory effects on C. vulgaris through excreting chemical substances [35] . Algicidal pigments produced by marine bacter ia have also been isolated and identified [45] . This is not an uncommon occurrence among co - cultures, as many bacteria are able to produce allelochemicals. Specifically, many gram - negative bacteria produce chemicals, such as acetylated homoserine lactones, that are used to regulate th e production of secondary metabolites and facilitate quorum sensing [45] . Although , the interaction between bacteria and algae is highly dynamic and the threshold for inhibition is often dependent upon culturing conditions, dominating microalgae strain, and resource availability [39] . T he ability of a bacteria l species to dominate over an algae species is also contingent with nutrient concentration and stability. The imbalance of nutrients may cause a normal ly symbiotic culture to transition into the collapse of an algal species through bacterial domination [38] . Additionally, the accumulation of toxins in recycled media may amplify these effects [34] . Another potential factor that should be considered while implementing recycled media is the use of anti - foaming agents. Anti - foaming agents are commonly employed in large - scale microalgae cul tivations in order to reduce foams occurring due to the introduction of gases into the culture medium [46] . Foaming is a potentially serious problem in bioreactors and can result in overflows, loss of culture and produc ts, along with operational problems with machinery such as pumps [47] . Antifoams may be composed of a variety of different materials, such as silicone or polypropylene glycol. It is known that certain antifoams can affect the growth rates and 14 surface properties of prokaryotic and eukaryotic organisms [46] . Both negative and positive effects have been shown, however, there is relatively little information on how antifoams affect biological processes [46] . One way in which antifoams may affect biological processes is through their effect on dissolved oxygen (DO) content and volumetric mass oxygen transfer coefficients within a system [4 6] . For example, Al - Masry showed that a silicone - based antifoam had negative impacts on the mass transfer coefficient and gas velocity within the culture media of 55 and 700 L airlift reactors [47] . In contrast, Koch et al showed that an antifoam containing silicone oil only had a significant effect at the beginning of a process but decreased over time, and had varying effects over the growth of the microbial strains tested [48] . It has been suggested that some organisms may possess the ability to utilize antifoams as a way to increase their growth rate and improve protein production [46] . However, the impact of various antifoaming agents on differing prokaryotic and eukaryotic organis ms still remains unknown. Effect on growth and composition The replenishment of freshwater in microalgae cultivation after harvesting is a common practice used to frequently purge unwanted contaminants, such as pathogenic bacteria and toxic compounds, that are thought to be detrimental to algae growth [11, 38] . The in troduction of recycled media into cultivation systems may contain microalgal cell debris and organic compounds, which may have negative impacts on biomass growth and composition [ 13] . However, certain strains of microalgae, such as C. sorokiniana , are known to be highly robust and tolerant to a wide range of conditions, allowing them to thrive even in harsh conditions that may develop from recycled media. Several studies have in vestigated the impact of recycled media on microalgae growth and composition over a variety of strains, culture conditions, and experimental scales. Studies show a 15 wide range of results from increasing to decreasing growth rate and differences in compositi on. A review performed by A. Shahid et al found that over a range of studies, Desmodesmus, Tetraselmis, Arthrospira, and Hormotila sp. generally showed the most promising growth on recycled media [13] . Some species, such as Scenedesmus sp. and Nannochloropsis salina, showed no negative impact on cell growth but had reduced protein and/or lipid composition [13] . Other strains, including Teradesmus obliquus and C. zofingiensis, showed increased lipid and carbohydrate content and improved biomass growth, respectively [13] . The review also discussed results based on whether or not recycled media was rep lenished with additional nutrients. For example, C. vulgaris was found to grow successfully in nutrient replenished media. Other species, including Scenedesmus sp. and C. kessleri, were found to grow for a finite number of cycles o n recycled , but not reple nished, media before negative impacts occurred [13] . Some of the reviewed studies found that the accumulation of organic matter, such as with Arthrospira platensis, were at fau lt for reduced growth [13, 38] . Fret et a l found that media recirculation had no impact on Nannochloropsis sp. and Tisochrysis utea productivity when cultivated on microfiltered replenished media [14] . Limited research is available regarding cultivation of C. s orokiniana on recycled media . However, a research study conducted by Spence found a 3 - 18% reduction in growth rate in C. sorokiniana cultured at lab scale when media was recycled and replenished 1 - 4 times, respectively [34] . SUMMARY OF KNOWLEDGE GAPS Through this literature review it can be shown that media recirculation in microalgae cultivation systems is a potential solution to minimize water and nutrient usage, thus minimizing cost and environmental impact. However, due to a lack of commercial systems, large - scale microalgae cultivation is still an immature technology and information on the effects of 16 recir culation is still limited [12] . Further research is required to determine the feasibility of large - scale media recirculation in photoautotrophic microa lgae cultivation. Specific research on microalgal strain C. sorokiniana within large - scale photobioreactor systems is also limited. This includes effects of recirculation on microbial community and C. sorokiniana , effects of recirculation on growth and bio mass composition, and the overall effect on system performance. Taking into account the reviewed literature, it is expected that accumulated compounds, both mineral and microbial, will pose a challenge to system stability of large - scale and long - term algae cultivation under recirculation condition s . OBJECTIVES AND HYPOTHESIS The overall hypothesis is that freshwater and nutrient utilization can be reduced through the integration and recirculation of alternative water sources and that recirculation would ha ve no effect on system stability over a finite period . Th e objectives of this research were to: 1) study system stability under recirculation conditions; 2 ) m inimize freshwater usage through media recirculation ; and 3 ) m inimize nutrient usage through media recirculation . 17 CHAPTER 2: CONTINUOUS MICROALGAE CULTIVATION ON FLUE GAS IN A WATER RECIRCULATING PHOTO - BIOREACTOR SYSTEM ABSTRACT Growth media rec irculation is a potential solution to address water and nutrient challenges in large - scale microalgae cultivation. A pilot - scale algae photobioreactor (APB) was used to culture C . sorokiniana on flue gas from the T.B. Simon Power Plant at Michigan State University. Two culture conditions, freshwater and recirculated boiler water, were conducted on the APB to determine the effect of recirculation on algal growth. T h e results showed that microalgae cultivation under recirculation conditions was stable over a period of four months. Biomass productivities during the 1 st through 4 th months of recirculation (0.26, 0.23, 0.20, and 0.18 g L - 1 d - 1 , respectively) were not significantly different than freshwater ( 0.2 2 g L - 1 d - 1 ). Furthermore, the relationship between eukaryoti c and bacterial domains remained consistent throughout the four months of recirculation ( 80.7, 87.1, 83.1, and 82.1%, respectively and 19.2, 12.8, 16.9 and 17.8%, respectively ). This was not significantly different than the abundance of each domain in fres hwater cultivation (8 3.7 % eukaryotic and 16 .2 % bacterial). A 1 m 3 photobioreactor was then envisioned for a mass, energy and exergy analysis. The mass balance analysis concluded that a 98% reduction in freshwater usage and 25% reduction in nutrients could be achieved during cultivation operating under recirculation conditions for 1 year, while maintaining a biomass productivity of 1.2 kg wet algal biomass and 0.4 kg CO 2 sequestered per day. The exergy balance analysis concluded that without considering sola r irradiation , the culture with water recirculation more than doubled the rational exergy efficiency of the freshwater culture. 18 INTRODUCTION Carbon dioxide ( ) is a major greenhouse gas and principal contributor of global climate change. A large source of e missions is derived from power generation , such as power plants, which accounted for approximately 38% of total energy - related CO 2 emissions in 201 8 [1] . While many methods for mitigation exist, biological sequestration through microalgae cultivation represents a possible solution to decrease atmospheric and mitigate global warming. Microalgae have high capabilities to capture and generate biomass, offering attractive advantages over other terrestrial sequestrat ion methods. They are able to duplicate cell biomass 100 times faster than terrestrial plants, capture 10 - 50 times more efficiently, and do not require arable land [2] . Certain strains of microalgae remain photosynthetically efficient under a large range of concentrations [3] . In addition to capture, microalgae biomass is characterized by high protein, lipid, and carbohydrate con tent, which provides a good source for value - added products including pharmaceuticals, biofuels, and nutritional supplements [3, 4] . As a result, microalgae cultivation offers an attractive solution for capture from major carb on - emitting sources such as power plants. Large - scale microalgae cultivations are typically performed under controlled conditions in either open or closed algae photobioreactor ( APB ) systems. Although o pen systems are characterized by lower capital and operating costs, closed APB s utilize CO 2 more efficiently while also offering higher biomass productivity [5, 7] . However, s everal major technical challenges on la rge - scale algal cultivation hinder commercial algal production, including lack of long - term stability, high water and nutrient requirements , and an extremely large area requirement to balance industrial CO 2 emissions [4, 5, 10] . Improvements in operational 19 efficiency within APB operations is essential for technical and economic feasibility of this technology on a commercial scale . One of the main operational limitations is cost and environmental impacts due to the consumption of freshwater and nutrients [11, 12, 14] . A constant supply of water and certain inorganic nutrients are required to achieve and maintain high biomass productivity [12] . However, freshwater and nutrient usage in microalgae cultivation pose challenges to system sustainability. Freshwater is a natural resource that is becoming increasingly scarce, as freshwater aquifers are currently facing unsustainable rates of extraction [12] . L ife cycle assessments (LCAs) find that the life cycle burden of microalgae cultivation comes from nutrient production occurring upstream of algae cultivation facilities [12] . Therefore, maximizing the efficiency of water and nutrient use is a critical factor in improving the feasibility of the technology. Growth medi a recycling is a potential solution to address water and nutrient challenges in large - scale microalgae cultivation [3, 11 - 14] . Recirculation c an reduce the freshwater requirement while also reducing nutrient usage. An LCA study performed by Yang et al found that when harvesting water was completely recycled, nutrient usage decreased by about 55% and freshwater usage decreased by 84% [12] . Fret et al achieved a 77% decrease in water and 68% reduction in nutrients using media recirculation [14] . However , t he full potential of media recirculation across many microalgal species has yet to be completely explored in the field of large - scale cultivation. This research studied a long - term media recirculation operation in a pilot - scale APB using the flue gas from a power plant . The objectives of the study were to minimize freshwater 20 and nutrient usage, evaluate biomass production and alga l assemblage stability of the cultivation under recirculation conditions , and compare the culture under media recirculation and freshwate r conditions using mass, energy, and exergy analyses . MATERIALS AND METHODS Algal Assemblage The microalgae strain Chlorella sorokiniana MSU was isolated from the Great Lakes region for use in seeding the APB. C. sorokiniana isolates were stored on Tris - Acetate - Phosphate (TAP) agar medium [49] at room temperature and exposed under constant fluorescent light. Modified liquid TAP medi um was used for all photoautotrophic cultures. The modified TAP medium is based on a reference study [50] and contained the followin g substances: 3.75 mmol L 1 of NH 4 NO 3 , 0.34 mmol L 1 of CaCl 2 2 O, 0.4 mmol L 1 of MgSO 4 2 O, 0.68 mmol L 1 of K 2 HPO 4 (anhydrous), 0.45 mmol L 1 of KH 2 PO 4 (anhydrous), and 0.09 mmol L - 1 FeCl 3 2 O. Nutrient stock solutions were prepared using deionized water. Microalga e cultivations were performed in the closed, but not aseptic, APB. Pilot photobioreactor system and operations A PHYCO 2 APB unit previously installed in the T.B. Simon Power Plant was used in this study [51] . All cultures were approximately 100 L. The culture was exposed to 24 h lighting conditions provided by twelve red and blue LED light bars (Independence LED Lighting LLC, USA) . The LED light bars delivered a continuous photo synthetic photon flux density (PPFD) of approximately 407 mol m - 2 s - 1 to support algae growth. PPFD was measured using a LI - 190R Quantum Sensor and LI - 250A light meter (LI - COR, Lincoln, Nebraska). Additional system specifics of the APB are provided elsewh ere [50] . The natural gas fired flue gas, containing 7.5 ± 1.15% v/v of CO 2 , was directly pumped from the stack into the APB at a flow rate of 120 21 L/m 3 /min. The unit ran for approximately 7 months. Data used in this study were collected from May 2 nd 2019 to November 15 th 2019. Tw o semi - continuous cultures were cultivated under t wo water sources (freshwater and recirculated boiler water). Initial harvesting began once biomass productivity reached 0.2 2 g L - 1 day - 1 . A 50% harvest ratio was previously optimized for maximizing biomass concentration and was thus the only harvesting ra tio used in this study [50] . The specified water source (freshwater or recirculated boiler water) was used to refill the APB reactor after harvesting . Water was stored in a 380 L storage tan k for up to 5 days before being fed into the APB. Boiler water was obtained directly from power plant boilers . Freshwater cultivation was performed using the tap water from Michigan State University . Under recirculation conditions, biomass was removed through centrifugation and remaining broth was recirculated back to the APB. Nutrients were replenished to that of modified TAP media, with the exception of total phosphorus, which was allowed to accumul ate within the reactor to simplify daily nutrient additions . T he pH was maintained at 6.6 ± 0.09 for the freshwater treatment and 6.2 ± 0.27 for the recycle treatment. Chemical analysis Samples were analyzed daily for dry biomass weight, pH and nutrient c oncentrations (total nitrogen (TN), total phosphorus (TP), nitrate (NO 3 - N) and ammonia (NH 3 - N )). Algal biomass was concentrated for biomass productivity measurement s using a Dolphin Alfa Laval MAB204 Centrifuge . Wet biomass was weighed and then dried at 10 5 C for 24 h for dry weight determination. Sample pH was measured using a pH meter (Fisherbrand accumet AB15 + Basic, Fisher Scientific Co., Pittsburgh, PA). N utrient concentrations were tested in the liquid supernatant using nutrient test kits (HACH Company, Loveland, Colorado) equivalent to EPA method s [52] . Trace element analysis of the liquid supernatant , elemental analysis, and biomass 22 composition of select sampl es was conducted by Dairy One (Ithaca, NY). Additional elemental (CHNS) analysis was conducted by Atlantic Microlab (Norcross, GA). Microbial analysis DNA Extraction Microbial community samples (1.5 mL) collected for DNA analysis were taken once per week t hroughout the study and stored at - 20 C until extraction. To remove nutrient media, algae sample s were centrifuged using an Eppendorf 5416R centrifuge at 10,000 rpm for 5 min and the supernatant was discarded. The remaining pellet was washed and resuspende d once with deionized water, and the supernatant was discarded. The final remaining pellet was used for DNA extraction using a DNeasy PowerSoil DNA Isolation Kit (Qiagen, Germany). DNA extracts were eluted with 100 L of 10 mM Tris - HCl (pH 8.5) and the c oncentration and purity were determined using a NanoDrop Lite spectrophotometer (Thermo Fisher Scientific, USA). Extracted DNA samples were stored at - 80 C for several weeks before use in real - time PCR quantification and high - throughput sequencing (Illumina MiSeq flow cell) . Illumina preparation and sequencing Illumina sequencing was performed for the 16S rRNA gene region to assess the bacterial community. The PCR conditions were as follows: 1.0 L DNA template (10x diluted of microbial community DNA ), 0.5 L of 100 M forward primer ( IDT, Pro341F 5 - CCTACGGGNBGCASCAG - 3 ), 0.5 L of 100 M reverse primer IDT, Pro805R 3 - GACTACNVGGGTATCTAATCC - 5 ), 12.5 L 2x Supermix (Invitrogen, USA), and 10.5 L PCR grade water. The PCR program used for all assays were as follows: 96 C for 2 min, followed by 30 cycles of 95 C for 20 s, 52 C for 30 s, and 72 C for 1 min, and a final elongation 23 period of 72 C for 10 min. Amplicons were quality - tested and size - selected using gel electrophoresis ( 1.0% (w/v) agarose concentration and 1× TAE run buffer). Samples were then diluted to normalize DNA concentrations within 5 - 10 ng L - 1 by measuring the DNA concentration with the PicoGreen dsDNA quantitation assay (Invitrogen, USA) and Fluostar Optima microplate rea der (BMG Labtech, Germany). T he normalized PCR products were then sequenced at the Michigan State University (MSU) Research Technology Support Facility (RTSF). Illumina MiSeq (pair - end 250 bp) targeting on V3_V4 hypervariable regions was used to carry out the sequencing. Fastq files from the high - throughput sequencing we re analyzed using the QIIME2 database to generate taxonomic/phylogenetic data for statistical analysis [53] . Statistical analysis All statistical analyses were performed using R statistical software (Versio n 3.6.3) . The data with normal distribution and equal variance w ere analyzed using one - way analysis of variance (ANOVA). When data violated the normality assumption and equal variance , the Kruskal - - wise rank comparison post - hoc tests were used following ANOVA and Kruskal - Wallis tests, respectively. A significance value of = 0.05 was used for all tests . Microbial analysis was performed us ing the R libraries V egan, ggplot2, phyloseq, and MASS on taxonomic/phylogenetic data to graph relative abundances of samples. Non - metric multi - dimensional scaling analysis (NMDS) was then used to correlate the dissimilarities between culture conditions, r eactor performance, and phenotype abundance. 24 Mass balance analysis A mass balance analysis was conducted on a 1 m 3 APB unit to compare freshwater and recirculation cultivation. The envisioned APB unit has a volume 10 times greater than the experimental te sting unit. The gas transfer in the APB is operated through airlift, which requires a high inlet flue gas flow rate of 212.5 kg per day. A 50% harvesting ratio is used for both scenarios. Flue gas CO 2 removal rate is based on biomass productivity and carbo n content of algal biomass under each condition. After separation of biomass from medi um using a centrifuge, medi um is discarded under freshwater conditions or reintroduced to the APB under recirculation conditions. Slightly overdosing nutrients is a commo n practice to prevent nutrient limitation [14] . Using this practice, any discharged media contain s a fraction of the initial nutrient content. All performance parameters are based on pilot - scale experimental data . Energy balance analysis An energy balance analysi s was conducted for the envisioned 1 m 3 APB unit based on the pilot operational data. The electricity input for the LEDs was based on an average photosynthetic photon flux density (PPFD) of 407 m - 2 s - 1 and reactor surface area of 46.2 m 2 . PPFD was converted to electrical input using the conversion factor of 1 w m - 2 to 2.1 m - 2 s - 1 [54] . The centrifuge is a 5 kW unit and the average running time is 3 min d - 1 . The harvesting wate r pump is a 0.5 kW unit with an average running time of 8 min d - 1 . The water refill pump is a 0.5 kW unit with an average running time of 8 min d - 1 . The inlet centrifuge pump is a 0.5 kW unit and runs for 3 min d - 1 . The outlet centrifuge pump is a 0.5 kW u nit and runs for 3 min d - 1 . The flue gas pump is a 0.146 kW unit with an average running time of 24 h d - 1 . The energy content of algal biomass as biofuel was not considered in the energy balance calculation. 25 Exergy analysis An exergy analysis was conducted on microalgae biomass accumulation under recirculation and freshwater conditions, based on the reference approach [55] . The mass and energy balance data were used to carry out the exergy analysis. The exergy flow rates of individual compounds were calculated as the sum of their physical and chemical exergy flow rates: where k is the k th component in the process, is the chemical exergy rate (kW) of the k th component, is the physical exergy rate (kW) of the k th component, and is the process exergy rate (kW) of the k th component. and are defined as follows: where is the mass flow rate (kg/day) of the k th component, is the specific chemical exergy (kJ/kg) of the k th component; and are the specific enthalpy (kJ/kg) and specific and are the spe cific enthalpy and specific entropy of the reference environment, respectively, and is the reference temperature (298.15 K). The physical exergies of components with similar temperature to the reference environment were negligible in comparison with ch emical exergy rates. The rational exergy efficiency ( ) was calculated as the desired exergy output divided by the used exergy input: 26 where is the exergy rate (kW) of the desired product i n the output stream and is the exergy (kW) used deduced from the exergies of wasted ( ) and desired ( ) products and irreversibility ( I ) of the system. and are defin ed as followed: where is the exergy rate (kW) of the de sired algal biomass product and is the exergy rate (kW) of the water in the desired product. is the irreversibility or exergy destruction (kW) of the process and is the sum of the exergy rates (kW) of the undesirable products in the output stream. and are defined as followed: where is the total exergy in the output stream (kW) and is the total exergy in the input stream (kW). The component calculation of and is expressed as followed: 27 The most significant exergy demand is from solar or illumination energy, which accounts for more than 9 4 % of the total daily exergy demand. As solar energy remains constant under water recirculation and freshwater conditions, rational exergy without solar energy was considered in order to delineate differences of lesser magnitude between two conditions. without solar irradiation ) was calculated as the d esired exergy output divided by the u sed exergy input without solar energy : where is the exergy (kW) used deduced from the exergies of wa sted ( ) and desired ( ) products and irreversibility ( ) of the system, without solar energy. is defined a s: where is the irreversibility (kW) of the process without solar energy , is defined as: 28 RESULTS AND DISCUSSION Effects of water recirculation on algae growth, nutrient consumption, and biomass composition The continuous cultures were conducted using the APB to i nvestigate the effects of water recirculation on algal biomass production, composition , and nutrient consumption. The data demonstrate that water recirculation does not impact biomass growth in terms of biomass productivity compared to freshwater condition . The data also suggest that while there were significant differences in nutrient consumption , these trends do not significantly impact total biomass production or composition between conditions. Monthly biomass productivity (Fig. 1a) and nutrient consump tion (Fig. 1b) were analyzed to determine the optimal recirculation period until recirculated water should be replaced with freshwater. Although biomass productivity showed a decreasing trend over a period of four months of water recirculation, the 1 st thr ough 4 th months of recirculation showed no significant differences (P > 0.05) in biomass productivity (0.26, 0.23, 0.20, and 0.18 g L - 1 d - 1 , respectively) in comparison to freshwater (0.22 g L - 1 d - 1 ). TN consumption during the 1 st and 2 nd month s of recircu lation (28.3 and 30.0 mg TN L - 1 d - 1 , respectively) was significantly different (P < 0.5) than freshwater (22.0 mg TN L - 1 d - 1 ), but was not significantly different (P > 0.05) during the 3 rd and 4 th months (20.6 and 22.9 mg TN L - 1 d - 1 , respectively) than freshwater . NH 3 - N consumption was significantly different (P < 0.05) than freshwater during the 2 nd , 3 rd and 4 th months of recirculation (9.1, 7.0 and 6.9 mg NH 3 - N L - 1 d - 1 , respectively) , but was not significant ly different (P > 0.05) th an freshwater (11.9 mg NH 3 - N L - 1 29 d - 1 ) during the 1 st month of recirculation (11.1 mg NH 3 - N L - 1 d - 1 ) . NO 3 - N consumption also showed significant differences (P > 0.05) during the 1 st , 2 nd and 4 th months (12.9, 11.7 and 11.3 mg NO 3 - N L - 1 d - 1 , respectively) , but was not significantly different (P > 0.05) than freshwater (6.1 mg NO 3 - N L - 1 d - 1 ) during the 3 rd month (7.41 mg NO 3 - N L - 1 d - 1 ) . TP consumption was significantly different (P < 0.05) during the 1 st and 3 rd months of recirculation (3.3 and 1.9 mg TP L - 1 d - 1 , respectively) but was not significantly different (P > 0.05) than freshwater (2.6 mg TP L - 1 d - 1 ) during the 2 nd and 4 th months of recirculation (2.8 and 2.2 mg TP L - 1 d - 1 , respectively). (a) (b) Figure 1 : Biomass production and nutrient consumption . a.) Biomass productivity over 4 - month recirculation period compared with freshwater. b.) Nutrient consumption over 4 - month recirculation period compared with freshwater. The biomass productivity data suggest that a four - month time period is an acceptable duration for water recirculation before recirculated water should be replaced with freshwater. Biomass productivity was not significantly different than freshwater during any month, regardless of differences in monthly n utrient consumption. Furthermore, biomass components ( both elements and macromolecules ) between recirculation and freshwater conditions showed similar content (Table 1) , suggesting that recirculation did not greatly affect biomass composition. However, fou r month s of recirculation is a conservative recirculation estimate, as 30 the 4 th month did not show significant differences in biomass productivity compared to the freshwater condition . Thus, the actual maximum recirculation period under the recirculation conditions cannot be concluded . Table 1 : Biomass composition under recirculation and freshwater conditions Component (% of biomass) Recy cle Freshwater C a 46.2 ± 0.89 46.4 ± 0.44 H a 7.1 ± 0.05 7.1 ± 0.08 O c 32.3 ± 1.08 30.5 ± 0.09 N a 8.5 ± 0.35 8.7 ± 0.27 P b 1.9 ± 0.10 2.4 ± 0.25 S a,b 0.6 ± 0.09 0.6 ± 0.09 Carbohydrate b , d 19.2 ± 2.61 19.5 ± 2.44 Lipid b 12.2 ± 0.47 9.7 ± 1.68 Protein b 59.1 ± 2.47 58.6 ± 2.41 Ash b 9.6 ± 0.36 12.2 ± 1.52 a: Value obtained from Atlantic Microlabs b: Value obtained from DairyOne Inc. c: Value obtained through calculating 100 - SUM (macro and micronutrient composition). No ash included. See appendix for details on detailed mineral content of biomass. d: Carbohydrates = NFC + NDF. See Table 4 in the appendix for composition values. Effects of water recirculatio n on algal assemblage Microbial community composition is a useful indicator for APB system operational status. The unique co - culture assemblage was monitored to further assess APB stability under recirculation conditions. Illumina library preparation and sequencing were performed at the MSU RTSF and . Sequence analysis showed a high abundance of cyanobacteria, which is known to have 85 - 93% 16s rRNA gene sequence similarity with microalgal strain C. sorokiniana [56] . Further analysis through microsc opic imaging did not indicate the presence of cyanobacteria, which are frequently characterized by 31 blue - green color and various shape, size and form, unlike C. sorokiniana [57] . Thus, the sequence was interpreted as microalga C. sorokiniana for all further analyses. The unique algae - bacteria assemblage was monitored to further assess APB stability under recirculation conditions. Microbes with less than 0.5% of relative abundance are not discussed. The relative abundances of the dominant microbial communities at three taxonomic levels (Domain, Phylum and Class) are presented in Fig. 2. Communities are shown over a four - month recirculation time period and under freshwater conditions. These communities belong to a total of 2 domains, 8 phyla and 9 classes. The domains present are e ukarya (containing microalgal species of interest, C. sorokiniana ) and b acteria (Fig. 2a). The dynamic data over the recirculation period show that the relationship between b acteria and algae abundance remained relatively stable. No significant differences (P > 0.05) were observed for eukarya abundance in the 1 st through 4 th m onth of recirculation (80.7, 87.1, 83.1, and 82.1%, respectively) in comparison to freshwater conditions (83.7%). Similarly, no differences (P > 0.05) in abundance were observed in the bacterial domain for the 1 st through 4 th month (19.2, 12.8, 16.9 and 17 .8%, respectively) compared to freshwater conditions (16.2%). 32 (a) (b) (c) (d) Figure 2 : Relative abundance of bacterial communities . a) at d omain level . b) at p hylum level . c) at Proteobacteria class level . d) at Bacteroidetes class level . At the bacterial phylum level, Proteobacteria (9.5 - 16.8%), Bacteroidetes (1.5 - 5.3%) and Bacteria unclassified (0 .3 - 2.5 %) abundance dominated both freshwater and recirculation communities (Fig. 2b). These findings are consistent with other environmental and large - scale microalgae cultivation studies [8, 58] . The phylum Proteobacteria (Fig. 2c) was the most ab undant group under both freshwater and recirculation conditions. Proteobacteria abundance in the 1 st month of recirculation (16.8%) was significantly higher (P < 0.05) than during freshwater conditions (9.5%) while the 2 nd , 3 rd and 4 th months of recirculat ion (10.1, 11.1 and 12.5%, respectively) showed no significant difference (P > 0.05). Under recirculation conditions, Proteobacteria unclassified was the most abundant (6.5 - 11.7%) class, followed by Alphaproteobacteria and Betaproteobacteria (2.0 - 4.2% 33 and 0.7 - 1.1%, respectively). Under freshwater conditions, Alphaproteobacteria and Betaproteobacteria were the most abundant classes (4.1 and 3.9%, respectively), followed by Proteobacteria unclassified (1.3%). No significant differences (P > 0.05) were obs erved for Alphaproteobacteria abundance during the 1 st through 4 th months of recirculation (4.2, 2.9, 2.9 and 2.0% , respectively ) when compared with freshwater conditions (4.1%). However, significant differences were identified in the abundance of the Beta proteobacteria and Proteobacteria unclassified classes. Betaproteobacteria abundance was significantly less (P < 0.05) during the 1 st through 3 rd months of recirculation (1.0, 0.7 and 1.1, respectively) compared to freshwater conditions (3.9%), although th e 4 th month showed no difference (0.9%). Proteobacteria unclassified abundance was significantly higher (P < 0.05) during all four months of recirculation (11.7, 6.5, 7.1 and 9.5%, respectively) compared to freshwater conditions (1.3%). The Bacteroidetes p hylum (Fig. 2d) was another dominant bacterial group among recirculation and freshwater. No significant differences (P > 0.05) were observed between any months of recirculation (1.5, 2.3, 5.3 and 4.9%, respectively) when compared to freshwater conditions ( 4.1 %) . However, significant differences (P < 0.05) were observed between the months of recirculation. Under both freshwater and recirculation conditions, the most abundant class within Bacteroidetes was [Saprospirae] (phylum is disputed) (1.1 - 4.4%). No sig nificant differences (P < 0.05) in [Saprospirae] were observed between the 1 st through 4 th months of recirculation (1.1, 1.9, 4.4 and 2.9%, respectively) when compared with freshwater abundance (2.8%). Although, significant differences (P < 0.05) were obse rved between months of recirculation. 34 (a) (b) Figure 3 : Non - metric multidimensional scaling ( NMDS) analysis of microbial community relative abundance for the APB system *. a.) Culture condition . b .) Water recirculation time . *: Green ellipses indicate culture condition and red indicate recirculation time. The red arrows indicate significant operational parameters (P < 0.05) of the APB. Blue arrows indicate significant microbial groups (P < 0.05). 35 N on - metric multidimensional scaling ( NMDS) analysis was conducted to determine the linkage of 16S rRNA gene sequences to environmental condition (Fig. 3a) and recirculation time (Fig. 3b). Experimental condition, significant operational parameters and key bacterial g roups were compared with microbial community distances to determine the interaction between microbial community, environmental condition, and water recirculation time. Recirculation conditions significantly (P=0.001) shifted the microbial communities (gree n ellipses in Fig. 3a), as did time (P=0.001) of recirculation (red ellipses in Fig. 3b). One bacterial phyla (Bacteroidetes) and three bacterial classes (Betaproteobacteria, [Saprospirae] and Proteobacteria unclassified) were significantly (P < 0.05) infl uenced under recirculation and freshwater conditions. Additionally, one bacterial phyla (Bacteroidetes) and two bacterial classes (Alphaproteobacteria and [ Saprospirae ] ) were significantly (P < 0.05) influenced by time of recirculation. Bacteroidetes was affected by both culture condition and recirculation time. Most notably, Fig. 3b shows the increased abundance of Bacteroidetes over increasing time of recirculation. Members of the Bacteroidetes phylum are highly diverse , but are frequently recognized as specialists for degrading complex organic matter, such as proteins and carbohydrates [58] . The accumulation of organic matter over the recirculation period, such as algal c ell components, is a potential carbon and energy source for Bacteroidetes. Unlike freshwater conditions, the accumulation of unassimilated compounds is a critical consideration for recirculated media. The class [Saprospirae], was found to be the most abun dant class within Bacteroidetes and was also significantly affected by culture condition and time. Recirculation time (Fig. 3b) showed a clear relationship with [Saprospirae], with relative abundances increasing over time. 36 Further analysis showed that Chit inophagaceae was the dominating bacterial family from the [Saprospirae] class. Previous studies suggest that Chitinophagaceae may provide molecules and enzymes which act as potential stimulants for plant growth [59] . Another study [60] on algal - bacteria symbiosis also reported the presence of Chitinophagaceae in association with algal growth in wastewater, and suggests that the growth of this bacterial group was favored due to its co - existence with algae. Fig. 3 suggests that significa nt changes in TN and NO 3 - N consumption (P=0.034 and P=0.001, respectively) occurred under recirculation conditions. This trend may indicate that the bacterial consortia utilized nitrogen compounds differently between freshwater and recirculation culture co nditions. Betaproteobacteria, for example, was found to be significantly influenced (P=0.001) by culture condition, showing a higher relative abundance under freshwater conditions than during recirculation (Fig. 3a). The Betaproteobacteria class is known t o contain many species of ammonia oxidation bacteria [61] , which provides a possible explanation for differences in nitrogen consumption patterns among culture conditions. A particularly interesting study conducted by Sambles et al discusses microbial community changes in an algae - bacteria co - culture due to repeated rinsing. Certain groups, such as Alphaproteobacteria, increased after rinsing [62] . Under recirculation conditions, Alphaproteobacteria was shown to decrease over time. It is possible that rinsing with freshwater may provide a favorable environment for this group o f bacteria. Sambles et al also identified certain bacterial orders that were not removed from the culture by rinsing , while others were removed after rinsing [62] . This suggest s that microbes not removed by rinsing were likely very closely associated or attached to algal colonies [62] . Bacteria unclassified and Proteobacteria unclassified were shown to be affected by culture condition and/or recirculation time. Although 37 these bacterial groups could not be further classified beyond the domain and phylum levels, respectively, speculations can be made as to why differences in abundance exist between cond itions. It is possible that daily harvesting disrupts algae - bacteria relationships. While this disruption remains permanent under freshwater conditions, media recirculation may reinforce such relationships, as previously rinsed microbes are eventually re - i ntroduced into the system. It is likely that more detailed relationships such as this are masked under currently unclassified groups. Furthermore, although biomass productivity under recirculation was not found to be significantly different than freshwater , it is clear in Fig. 3b that biomass productivity showed an overall decreasing trend over the recirculation period. These results are consistent with similar small - scale algae cultivation studies on media recirculation in microalgae cultivation [14] . It remains u ncertain why biomass decreased over recirculation. One potential cause for decrease may include unidentified pathogenic bacteria that were allowed to grow under prolonged residence time within the system and harm algal cells. Some species in the order of C ytophagia, for example, are capable of lysing a variety of algae cells [8] . Another possible facto r is the gradual accumulation of substances that may be detrimental to algal health. Potential accumulating substances could be inhibitory compounds from bacterial or algal cells, nutrients, minerals, or antifoaming agent used to control foaming during alg ae cultivation. Compounds such as inhibitory substances and antifoaming agent are not assimilated into the microalgae biomass, allowing concentration to rise as media is recirculated. Select nutrients and minerals, such as potassium (K) and P were also fou nd to accumulate within the reactor, as algal biomass consumption rate was likely less than the amount added each day. Table 8 in the appendix 38 provides more detailed information on micronutrient accumulation in the culture media over the recirculation peri od and under freshwater conditions. Microbial analysis demonstrated that the relationship between microalgae and bacteria domains was relatively stable for the recirculation period and did not display and major differences between conditions. However, it i s apparent that the intricacies and complexities of microbial community on biomass growth cannot be fully characterized by the obtained results. Understanding microbial community structure and function within microalgae cultivation is essential for large - s cale microalgae cultivation. In order to determine the true stability and functional relationships between microalgae and bacteria, more detailed analysis on community structure and function at genus or species level is required. Mass and energy balance of the water recirculating photobioreactor system Using experimental data, a mass and energy balance analysis was conducted on a 1 m 3 A PB unit (Fig. 4 and Table 3). Under freshwater conditions, the 1 m 3 unit produces 1.1 kg of wet algal biomass per day (Fig. 4a). The daily nutrient requirements are 0.036 kg nitrogen and 0.008 kg of phosphorus. Additionally, 500 kg of water is required each day to replenish 498.9 kg of water discharged from centrifuging and to replace water contained in algal biomass. The discharged water contains approximately 0.011 kg nitrogen and 0.003 phosphorus. The culture sequesters approximately 0.4 kg of CO 2 per day . 39 (a) (b) Figure 4 : Mass balance of the pilot APB unit *. a.) Mass balance under freshwater conditions . b .) Mass balance under recirculation conditions. * : Calculation of CO 2 fixation under freshwater conditions is based on microalgal cell formula: CH 1.84 O 0.49 N 0.16 P 0.020 S 0.005 Fe 0.009 K 0.008 Ca 0.005 Mg 0.003 *: Calculation of CO 2 fixation under recirculation is based on microalgal cell formula: CH 1.84 O 0.52 N 0.16 P 0.016 S 0.005 Fe 0.006 K 0.009 Ca 0.003 Mg 0.003 Under recirculation conditions, the photobioreactor produces 1.2 kg of wet algal biomass per day (Fig. 4b). The nutrients required are 0.027 kg nitrogen and 0.006 kg of phosphorus per day, and 495.8 kg of water is able to recycle after separation of algal biomass from media by centrifugation. Recirculated water contains 0.005 kg of nitrogen and 0.011 kg of phosphorus. Thus, only 4.2 kg of freshwater is required each day to replenish 3 kg of water discharged from centrifuging and to replace water contained i n algal biomass. The culture sequesters approximately 0.4 kg of CO 2 per day. 40 The mass balance analysis shows that for a 1 - year cultivation operating under recirculation conditions, approximately 179,000 kg of freshwater, 3.2 kg of nitrogen, and 0.7 kg of p hosphorus can be conserved (Table 2). Th is represent s a 98% reduction in freshwater, and 25% reduction of both nitrogen and phosphorus over freshwater cultivation. However, it should be noted that because phosphorus was allowed to accumulate to higher concentrations under recirculation than in freshwater con ditions, the theoretical P requirement to maintain the same biomass productivity and CO 2 sequestration rate may be lower than reported. Table 2 : Resources saved of the envisioned 1000 L algae photobioreactor for 1 - year operation under recirculation condi tions * . Resource Amount Saved (kg) Reduction Freshwater 179,000 98 % Nitrogen 3.2 25 % Phosphorus 0.7 25 % *4 - month water usage before replacing with new freshwater * B ased on mass balance and pilot operational data * In comparison with freshwater treatment conditions The energy balance analysis shows that a net energy input of 219 kWh - e unit - 1 day - 1 is required to power the APB unit under both freshwater and recirculation conditions (Table 3). The electricity demands for the LED ligh ts and centrifuge are 215 and 0.25 kWh unit - 1 day - 1 , respectively. The energy required for the APB water addition and culture harvesting pumps are 0.07 and 0.07 kWh unit - 1 day - 1 , respectively. The energy required for centrifugation pumps is 0 . 03 kWh unit - 1 day - 1 , eac h . The energy requirement for the flue gas pump is 3.50 kWh unit - 1 day - 1 . The most significant energy demand comes from the electrical input to the LED lights , which accounts fo r 98 % of the total daily energy demand. 41 Table 3 : Energy balance of the envisioned 1000 L algae photobioreactor * Component Energy Required Electrical input to LED lights ( k Wh/unit/day) - 215 Electrical input to centrifuge ( k Wh/unit/day) - 0.25 Electrical input to water addition (kWh/unit/day) - 0.07 Electrical input to culture harvesting (kWh/unit/day) - 0.07 Electrical input to centrifuge pump (tank to centrifuge) (kWh/unit/day) - 0.0 3 Energy input to centrifuge pump (centrifuge to sink) (kWh/u nit/day) - 0.0 3 Energy input to flue gas pump (kWh/unit/day) - 3.50 Net energy (kWh - e/unit/day) - 21 9 * : The energy balance analysis was based on the mass balance and pilot operational data. Energy input is negative and energy output is positive. Exergy analysis of the water recirculating photobioreactor system An exergy analysis was conducted to calculate the rational exergy efficiencies and compare process effectiveness between water recirculation and freshwater conditions (Fig. 5). The rational exergy efficiency under recirculation and freshwater conditions were 0.6 4 and 0.59%, respectively. The rational exergy efficiencies without considering solar energy were 2 3 and 10%, for recirculation and freshwater, respectively. The data indicate that microalgae cultivation under recirculation conditions had a better exergetic performance than cultivation under fr eshwater conditions. Figure 5 : Rational exergy efficiencies of freshwater and water recirculation cultivations. 0 5 10 15 20 25 without solar energy Exergy efficiency (%) Cutlivation with boiler water recirculation Cultivation with fresh water 42 The freshwater control used a greater amount of chemicals and freshwater in the input stream and wasted a greater amount of product in the o utput stream, while maintaining a similar output in the desired product compared to recirculation. In contrast, recirculation minimized water and nutrient usage in the input and output streams, as water loss from centrifugation and flue gas were the only o utput flow s with undesirable products. Furthermore, the input stream required less water and nutrient input per day due to cross - over products from recirculation. Microalgae cultivation under recirculation conditions clearly exhibits an improved exergetic performance over freshwater conditions. The detailed exergy analysis is provided in the appendix (Table s 16 and 17 ). CONCLUSION A pilot - scale APB system for flue gas CO 2 sequestration and algal biomass production for microalgae species C. sorokiniana was s tudied under boiler water recirculation and freshwater conditions. The results indicate that water recirculation does not affect C. sorokiniana cultivation over the studied recirculation period. Biomass productivity was not significantly different between recirculation and freshwater conditions ( 0.23 and 0.2 2 g L - 1 d - 1 , respectively) and biomass content between conditions was comparable. Furthermore, the relationship between eukaryotic and bacterial domains remained stable between recirculation (81%, 87%, 83%, and 82%, respectively and 19%, 13%, 17% and 18%, respectively) and freshwater (84% eukaryotic and 16% bacterial). The water footprint was substantially reduced by 98% and nitrogen and phosphorus were each reduced by 25%. Recir culation conditions exhibited a n increase in rational exergy efficiency more than double tha t of freshwater conditions. 43 ACKNOWLEDGEMENTS The authors would like to thank the U.S. Department of Energy for their financial support ( grant # DE - FE0030977). The authors would like to acknowledge the assistance and support from Mr. Nathan Verhanovitz at the MSU T.B. Simon Power Plant and Sibel Uludag - Demirer, Ashley Cutshaw, Henry Frost, Yurui Zheng and Annaliese Marks, for the assistance in APB operations and rout ine chemical analysis. 44 CHAPTER 3. CONCLUSIONS AND FUTURE WORK CONCLUSIONS C. sorokiniana is a species of green microalgae that is commonly employed in large - scale cultivations for CO 2 sequestration . In order to improve the economic efficiency and susta inability of microalgae cultivation operations, fresh water and nutrient usage need to be reduced. Thus, C. sorokiniana was cultivated on flue gas CO 2 under boiler water recirculation and freshwater conditions to determine the feasibility of media recirculation in a pilot - scale operation. This research shows that the cultivation of C. sorokiniana on recycled media did not affect the growth in comparison to freshwater conditions in the pilot - scale unit over the recirculation period. No significant differences were observed in biomass productivity, composition, CO 2 sequestration and dominating micro bial domains. Meanwhile, m ass and exergy balances found that substantial amount s of nutrients and water could be conserved , therefore improving the overall efficiency of the system. It is important to note that a decreasing trend in biomass productivity wa s observed during the recirculation period and that s ignificant differences were observed in parameters including nutrient consumption and microbial community at phylum and class level . The accumulation of unassimilated compounds was also observed. However , the underlying cause behind microalgal biomass productivity decrease was not identified in this study and therefore remains uncertain. Potential contributors to this trend are increased abundance of competing microorganisms, pathogenic bacteria, accumula tion of unnecessary compounds, and likely, a combination of these. Overall, this study concluded that media recirculation is a feasible approach in pilot and large - scale microalgae cultivation systems over the finite period of four months . After this fini te 45 period, it is recommended that recirculated media should be replaced with freshwater to avoid significant decrease in biomass productivity . FUTURE WORK Future work in the area of media recirculation should be conducted to improve large - scale microalgae cultivation. The effect of water recirculation on additional economically relevant microalgal strains should be investigated, as various strains and microbial communities differ in response to cultivation condition s . Furthermore, a dynamic assessment of microbial community over the recirculation period should be performed. Closely monitoring changes in microbial community over time is crucial for identifying fluctuations in nutrient consumption and algal biomass growth. It also highlights a reas for potential optimization, such as reducing bacterial communities that compete with microalgae for resources or enhancing communities that are beneficial to microalgal growth. Additionally, accumulati ng and inhibitory compounds under recirculation co nditions should be identified and further characterized for their relationship with biomass productivity. The characterization of these compounds in combination with identification of harmful microbial groups can be used t o recognize additional measures th at should be taken to maximize recirculation period, such as media filtration or sterilization. The listed studies will contribute broadly to the area of microalgae cultivation by providing insight on areas that require optimization, thus further reducing natural resource consumption and increasing economic and environmental viability. 46 APPENDI CES 47 APPENDI X A: TABLES AND FIGURES ( a) ( b) ( c) Figure 6 : 100 L photobioreactor set up . a.) Helical coil (front) and up - tube (back). b.) Helical coil. c.) Red and blue LED light strip. Figure 7 : C. sorokiniana and microbial community. 48 Table 4 : C. sorokiniana biomass composition . * *: Biomass processed by DairyOne Inc. 49 Table 5 : C. sorokiniana CHNS composition . * *: Data from Atlantic Microlab . Table 6 : C. sorokiniana fatty acid composition as % of dry matter . * * : Results from DairyOne, Inc. 50 Table 7 : C. sorokiniana fatty acid composition as % of total fatty acids . * * : Results from DairyOne, Inc. Table 8 : Algal supernatant and boiler water composition . * *Result from DairyOne, Inc. 51 Table 9 : Biomass data under recirculation conditions . 52 Table 10 : Biomass data under freshwater conditions . 53 Table 11 : Operational data for recirculation cultivation . 54 Table 11 ) 55 Table 11 ) 56 Table 12 : Operational data for freshwater process. 57 Table 12 ) 58 Table 13 : Illumina OTU classification key . * *: _ [ ] means the OTU is unclassified . 59 Table 14 : Illumina sequencing abundance under recirculation conditions . * *: See Table 13 for OTU identification key. 60 Table 15 : Illumina sequencing abundance under freshwater conditions . * *: See Table 13 for OTU identification key. 61 Figure 8 : Overall microbial community for freshwater and recirculation samples. ( a) ( b) Figure 9 : M icrobial community abundance a.) at phylum level. b.) at class level. 62 ( a) ( b) ( c) Figure 10 : Heatmaps of m icrobial community abundance a.) at bacterial phylum b.) within Bacteroidetes phylum and c.) within Proteobacteria phylum . ( a) ( b) Figure 1 1 : Rarefaction curve s a.) R ecirculation and freshwater conditions and b.) I ndividual recirculation (S8 - 22) and freshwater (S23 - 35) samples. 63 ( a) ( b) ( c) ( d) Figure 1 2 : Scatter plots for community indices a.) b.) c.) d.) 64 ( a) ( b ) ( c ) ( d ) Figure 1 3 : Boxplots for microbial community indices comparing recirculation and freshwater conditions a.) b.) c.) d.) Index Figure 1 4 : Dendrogram of individual recirculation (S8 - 22) and freshwater (S23 - 35) sa mples. 65 Table 16 : Process compounds, mass flow rate, temperature, chemical exergy rate, physical exergy rate, and total exergy rate of individual compounds for microalgae cultivation under recirculation conditions . * - considered in the exergy efficiency calculation . 66 Table 17 : Process compounds, mass flow rate, temperature, chemical exergy rate, physical exergy rate, and total exergy rate of individual compounds for microalgae cultivation under freshwater condi tions . * - 67 APPENDIX B: R CODE FOR PLOTTING AND ANALYSIS N on - metric multidimensional scaling ( condition ) ## NMDS analysis for Microalgae Cultivation (Lumped) ## Wei Liao, March 10, 2020 ## Carly Daiek, March 20, 2020 update # Loading Library and Tables ---------------- # Load "vegan" and "MASS" libraries in R library(vegan) library(MASS) # Load data files, make sure the data files are saved as macintosh .csv and fo llow the sample format species < - read.csv(file.choose(), head = TRUE, row.names = 1) env < - read.csv(file.choose(), head = TRUE, row.names = 1) performance < - read.csv(file.choose(), head= TRUE, row.names = 1) #Statistical analysis --------------------- species.mds < - metaMDS(species, trace=FALSE) ef.sp < - envfit(species.mds, env, permu=999) perf.sp < - envfit(species.mds, performance, permu=999) species.mds ef.sp perf.sp # Plotting NMDS chart ------------------- plot(species.mds, display="sites", type="points") with(env, ordiellipse(species.mds, Recycle, kind= "se", draw="polygon", col="green", alpha=50, label=TRUE,border=NA, conf=0.95)) # With significant performance data ef.perf < - envfit(species.mds, performance[,c(2,5)], permu =999) plot(ef.perf, col="red", cex=0.8) # With significant microbial community data ef.perf < - envfit(species.mds, performance[, c(9,1 6 ,17, 20 )], permu=999) plot(ef.perf, col="blue", cex=0.8) N on - metric multidimensional scaling ( recirculation dynamics ) ## NMDS analysis for Microalgae Cultivation (Dynamic) ## Wei Liao, March 10, 2020 ## Carly Daiek, March 20, 2020 update # Loading Library and Tables ---------------- # Load "vegan" and "MASS" libraries in R library(vegan) library(MASS) 68 # Load data files, make sure the data files are saved as macintosh .csv and follow the sample format species < - read.csv(file.choose(), head = TRUE, row.names = 1) env < - read.csv(file.choose(), head = TRUE, row.names = 1) performance < - read.csv(file.choose(), head= TRUE, row.names = 1) # Statistical analysis --------------------- species.mds < - metaMDS(species, trace=FALSE) ef.sp < - envfit(species.mds, env, permu=999) perf.sp < - envfit(species.mds, performance, permu=999) species.mds ef .sp perf.sp # Plotting NMDS chart ------------------- plot(species.mds, display="sites", type="points", xlim=c( - 1 , 1 )) with(env, ordiellipse(species.mds, Month, kind= "se", draw="polygon", col="darkred", alpha=50, label=TRUE,border=NA, conf=0.95)) # With significant performance data ef.perf < - envfit(species.mds, performance[, c(1)], permu=999) plot(ef.perf, col="red", cex=0.8) # With significant microbial community data ef.perf < - envfit(species.mds, performance[, c(9,1 5 ,2 0 )], permu=999) plot(ef.perf, co l="blue", cex=0.8) B iomass statistics and plotting ## Algal Cultivation: Dynamic analysis of Recirculation vs. Freshwater ## BIOMASS ## Wei Liao ## Carly Daiek, February 2020 update # Loading Library and Tables ---------------------------------------------- library (MASS) library(ggplot2) library(grid) library(gridExtra) library(ggpubr) library(plyr) library(RVAideMemoire) library(DescTools) library(PMCMRplus) library(inferr) # Installing the font package ----- ---------------------------------------- library(extrafont) font_import() #It may take a few minutes to import. loadfonts(device="win") # PROGRAM TO PLOT BAR CHART WITH STANDARD DEVIATION ----------------------- #+++++++++++++++++++++++++ # Function to c alculate the mean and the standard deviation 69 # for each group #+++++++++++++++++++++++++ # data : a data frame # varname : the name of a column containing the variable #to be summarized # groupnames : vector of column names to be used as # grouping variables data_summary < - function(data, varname, groupnames){ require(plyr) summary_func < - function(x, col){ c(mean = mean(x[[col]], na.rm=TRUE), sd = sd(x[[col]], na.rm=TRUE)) } data_sum< - ddply(data, groupnames, .fun=summary_func, varname) data_sum < - rename(data_sum, c("mean" = varname)) return(data_sum) } # ANALYSIS --------------------------------------------------------------- ## the .txt file needs to be saved as the type of "Tab delimited". ##choose the meta data_biomass_dynamic, should be .txt con < - file.choose(new = FALSE) mastermetadata < - read.table(con, header = T, row.names = 1,na.strings=c("","NA"," "," ")) metadata < - mastermetadata metadata$month < - factor(metadata$month) # STEP 1 Assumption 1: norm ality of each group #Biomass Productivity byf.hist(biomass_productivity~month, density=TRUE, sep=FALSE, data = metadata) byf.shapiro(biomass_productivity~month, data = metadata) #Shapiro - Wilk Test #All months are normal # STEP 2 Checking assumption 2: variance #Biomass Productivity infer_levene_test(data=metadata, biomass_productivity, group_var = "month") #Levene F test #All months have equal variance ## STEP 3a Run a regular ANOVA if BOTH assumptions met ANOVA_BiomassProductivty < - aov (biomass_productivity~month, data=metadata) summary (ANOVA_BiomassProductivty) #Some or all means signficantly different if P < 0.05 # residual plots to spot unequal variance, lack of normality of residuals, & outliers with (metadata, par (mfrow=c(2,2))) p lot (aov(biomass_productivity~month, data=metadata)) #At least one month is significantly different ## Step 3b IFF your ANOVA was significant, run a post hoc ## Tukey's HSD for ALL pairwise TukeyHSD(ANOVA_BiomassProductivty) # PLOTTING ----------------- ---------------------------------------------- ## ORGANIZE DATA FOR PLOTTING # Biomass productivity BiomassProductivity < - data_summary(metadata, varname="biomass_productivity", 70 groupnames=c("month")) BiomassProductivi ty$month=as.factor(BiomassProductivity$month) ## PLOTTING #Grouped bar plot (Nutrient reduction by Month) values < - BiomassProductivity[,2] sd < - BiomassProductivity[,3] condition < - c("Freshwater", "1", "2", "3", "4") df < - data.frame(values, condition) box_1 < - ggplot(df, aes(x=factor(condition), y=values)) + geom_bar(stat="identity", position=position_dodge(), colour="black")+ geom_errorbar(aes(ymin=values - sd, ymax=values+sd), width=0.2, position=position_dodge(0.9))+ ylab("Biomass Pro ductivity (g/L/day)") + ylim(0, 0.4)+ labs(title = "", subtitle=NULL) + xlab ("Month") + theme(title=element_text(size=20, family="Times New Roman"), axis.text.x = element_text(size=16, family="Times New Roman"), axis.text.y=element_t ext(size=16, family="Times New Roman"), axis.title.y = element_text(size = 20, family="Times New Roman"), axis.title.x=element_text(size=20, family="Times New Roman"), legend.title=element_blank(), legend.text = element_te xt(size = 14, family="Times New Roman"), legend.position = "top") box_1 O perational statistics and plotting ## Algal Cultivation: Dynamic analysis of Recirculation vs. Freshwater ## Operational ## Wei Liao ## Carly Daiek, February 2020 update # Loading Library and Tables ---------------------------------------------- library(MASS) library(ggplot2) library(grid) library(gridExtra) library(ggpubr) library(plyr) library(RVAideMemoire) library(DescTools) library(PMCMRplus) library(inferr) # Installing the font package --------------------------------------------- library(extrafont) font_import() #It may take a few minutes to import. loadfonts(device="win") # ANALYSIS ------------------------------------- -------------------------- ## the .txt file needs to be saved as the type of "Tab delimited". ##choose the metadata_operational_dynamic, should be .txt con < - file.choose(new = FALSE) mastermetadata < - read.table(con, header = T, row.names = 1,na.strings=c( "","NA"," "," ")) metadata < - mastermetadata 71 metadata$month < - factor(metadata$month) ## STEP 1 Assumption 1: normality #TN CONSUMED: all months normal byf.hist(TN_reduction~month, density=TRUE, sep=FALSE, data = metadata) byf.shapiro(TN_reduction~mont h, data = metadata) #Shapiro - Wilk Test #TP CONSUMED: 1st month non - normal, all other months normal byf.hist(TP_reduction~month, density=TRUE, sep=FALSE, data = metadata) byf.shapiro(TP_reduction~month, data = metadata) #NH3 CONSUMED: Freshwater non - normal, all other months normal byf.hist(NH3_N_reduction~month, density=TRUE, sep=FALSE, data = metadata) byf.shapiro(NH3_N_reduction~month, data = metadata) #NO3 CONSUMED: Freshwater non - normal, all other months normal byf.hist(NO3_N_red uction~month, density=TRUE, sep=FALSE, data = metadata) byf.shapiro(NO3_N_reduction~month, data = metadata) ## STEP 2 Assumption 2:Variance #TN CONSUMED: unequal variance infer_levene_test(data=metadata, TN_reduction, group_var = "month") #Levene F tes t #TP CONSUMED: unequal variance infer_levene_test(data=metadata, TP_reduction, group_var = "month") #NH3 CONSUMED: equal variance infer_levene_test(data=metadata, NH3_N_reduction, group_var = "month") #NO3 CONSUMED: unequal variance infer_levene_test(da ta=metadata, NO3_N_reduction, group_var = "month") ## STEP 3 Non - Parametric Alternative to ANOVA ## KRUSKAL - WALLIS TEST if normality tests fail AND transforms cannot fix kruskal.test(TN_reduction~month, data=metadata) #KW test -- > shows signficant differe nce kruskal.test(TP_reduction~month, data=metadata) #KW test -- > shows signficant difference kruskal.test(NH3_N_reduction~month, data=metadata) #KW test -- > shows significant difference kruskal.test(NO3_N_reduction~month, data=metadata) #KW test -- > shows significant difference #####GRAPH BOX PLOT (USE W/ KW TEST) # KW indirectly compares medians, use box plots to visualize ggplot(metadata, aes(x = month, y = TN_reduction, fill = month)) + geom_boxplot() ggplot(metadata, aes (x = month, y = TP_reduction, fill = month)) + geom_boxplot() ggplot(metadata, aes(x = month, y = NH3_N_reduction, fill = month)) + geom_boxplot() ggplot(metadata, aes(x = month, y = NO3_N_reduction, fill = month)) + geom_boxplot() ## Post Hoc Tests fo r KW ## Conover Test kwAllPairsConoverTest(TN_reduction~month, p.adjust="bonf", data=metadata) #all pairwise kwAllPairsConoverTest(TP_reduction~month, p.adjust="bonf", data=metadata) #all pairwise kwAllPairsConoverTest(NH3_N_reduction~month, p.adjust="bonf ", data=metadata) #all pairwise kwAllPairsConoverTest(NO3_N_reduction~month, p.adjust="bonf", data=metadata) #all pairwise # PROGRAM TO PLOT BAR CHART WITH STANDARD DEVIATION ----------------------- #+++++++++++++++++++++++++ # Function to calculate the m ean and the standard deviation # for each group #+++++++++++++++++++++++++ # data : a data frame # varname : the name of a column containing the variable #to be summarized # groupnames : vector of column names to be used as 72 # grouping variables data_summar y < - function(data, varname, groupnames){ require(plyr) summary_func < - function(x, col){ c(mean = mean(x[[col]], na.rm=TRUE), sd = sd(x[[col]], na.rm=TRUE)) } data_sum< - ddply(data, groupnames, .fun=summary_func, varname) data_sum < - rename(data_sum, c("mean" = varname)) return(data_sum) } # PLOTTING --------------------------------------------------------------- # Data sorting data1 < - metadata #ORGANIZE DATA FOR PLOTTING # CONSUMPTION DATA # TN reduction TNreduction < - data_summary(metadata, varname="TN_reduction", groupnames=c("month")) TNreduction$month=as.factor(TNreduction$month) # TP reduction TPreduction < - data_summary(metadata, varname="TP_reduction", groupnames=c("month")) TPreduction$month=as.factor(TPreduction$month) # NH3 reduction NH3reduction < - data_summary(metadata, varname="NH3_N_reduction", groupnames=c("month")) NH3reduction$month=as.fac tor(NH3reduction$month) # NO3 reduction NO3reduction < - data_summary(metadata, varname="NO3_N_reduction", groupnames=c("month")) NO3reduction$month=as.factor(NO3reduction$month) ## PLOTTING #Grouped bar plot (Nutrient reduction by Month) values1 < - c(TNreduction[,2],TPreduction[,2], NH3reduction[,2],NO3reduction[,2]) sd1 < - c(TNreduction[,3],TPreduction[,3], NH3reduction[,3],NO3reduction[,3]) condition1 < - rep(c("Freshwater", "1", "2", "3", "4") , 4) nutrient 1 < - c(rep("TN",5), rep("TP",5), rep("NH3 - N",5), rep("NO3 - N",5)) df1 < - data.frame(values1, condition1, nutrient1) box_1 < - ggplot(df1, aes(x=factor(condition1), y=values1, fill=nutrient1)) + geom_bar(stat="identity", position=position_dodge(), colour=" black")+ geom_errorbar(aes(ymin=values1 - sd1, ymax=values1+sd1), width=0.2, position=position_dodge(0.9))+ ylab("Nutrient Consumed (mg/L/day)") + ylim(0, 40) + labs(title = "", subtitle=NULL) + xlab("Month")+ theme(title=element_text(size=20, famil y="Times New Roman"), axis.text.x = element_text(size=16, family="Times New Roman"), axis.text.y=element_text(size=16, family="Times New Roman"), axis.title.y = element_text(size = 20, family="Times New Roman"), axis.tit le.x=element_text(size=20, family="Times New Roman"), legend.title=element_blank(), legend.text = element_text(size = 16, family="Times New Roman"))+ 73 scale_fill_manual(values=c("#999999", "#E69F00", "#56B4E9", "111111")) box_1 Part A (co ndition) ## Metagenomic analysis ## Algal Cultivation: Lumped analysis of Recirculation vs. Freshwater ## Part A ## Wei Liao ## Carly Daiek, February 2020 # Loading Library and Tables ---------------------------------------------- library(vegan) library(phyloseq) library(MASS) library(ggplot2) library(grid) library(gridExtra) library(ggpubr) ## the .txt file needs to be saved as the type of "Tab delimited". ## Gene frequency data from QIIME2 ##Choose the Frequency_Table_average should be a .txt con < - file.choose(new = FALSE) ##Now choose the Frequency_Table_Taxanomy should be .txt con1 < - file.choose(new = FALSE) ## Now we create the data.frame used for Frequency Table. Frequency_Tabl e < - read.table(con, header = T, row.names = 1) Frequency_Table_taxonomy < - read.delim(con1, header = T, row.names = 1) ## Alpha Diversity --------------------------------------------------------- t.Frequency.table < - t(Frequency_Table) # Transpose the d ata class(t.Frequency.table) # Check the class of the table #Rarefaction ----------------------------------------------- col < - c("black", "darkred", "forestgreen", "orange", "blue", "yellow", "hotpink") lty < - c("solid", "dashed", "longdash", "dotdash") pars < - expand.grid(col = col, lty = lty, stringsAsFactors = FALSE) ra < - rarecurve(t.Frequency.table, step = 20, col =col,lty = lty, cex = 0.6) # Rarefaction Curve Part A ( recirculation dynamics ) ## Metagenomic analysis ## Algal Cultivation: Dynamic analysis of Recirculation vs. Freshwater ## Part A ## Wei Liao ## Carly Daiek, February 2020 # Loading Library and Tables ---------------------------------------------- library(vegan) library(phyloseq) l ibrary (MASS) 74 library(ggplot2) library(grid) library(gridExtra) library(ggpubr) ## Gene frequency data from QIIME2 # Installing the font package --------------------------------------------- library(extrafont) font_import() #It may take a few minutes to import. loadfonts(device="win") #IMPORT DATA ## the .txt file needs to be saved as the type of "Tab delimited". ##Choose the Frequency_Table should be a .txt con < - file.choose(new = FALSE) ##Now choose the Frequency_Table_Taxonomy should be .txt con1 < - file.choose(new = FALSE) Frequency_Table < - read.table(con, header = T, row.names = 1) Frequency_Table_taxonomy < - read.delim(con1, header = T, row.names = 1) # Alpha Diversity ------------------------------ --------------------------- ## Now we create the data.frame used for Frequency Table. ## Now we create a matrix object with the data frame t.Frequency.table < - t(Frequency_Table) # Transpose the data class(t.Frequency.table) # Check the class of the tab le # Alpha diversity analysis indexes #First Shannon H < - diversity(t.Frequency.table, index = "shannon", MARGIN = 1, base = exp(1)) #Then Simpson D < - diversity(t.Frequency.table, "simpson", MARGIN = 1, base = exp(1)) #Third inverse Simpson iD < - divers ity(t.Frequency.table, "inv") # The last is Pielou's evenness J< - H/log(specnumber(t.Frequency.table)) ##List all indexes IN < - cbind(H,D,iD,J) write.csv(IN, "diversity.csv") ##Let's plot H, D, iD, and J par(mfrow=c(2,2)) plot(H) plot(D) plot(iD) plot(J) ## Estimate Chao1 and ACE estimateR(t.Frequency.table) # ANOVA for Alpha Diversity ---------------------------------------------- # using the H, D, iD, and J data to generate "alphadiversity.txt" to run one way ANOVA # choose the alpha diversity data should be .txt (tab delimited) 75 con3 < - file.choose(new = FALSE) alphadiversity < - read.table(con3, header = T, row.names = 1) alphadiversity$recycle < - factor(alphadiversity$recycle) ##Factor Statement #ANOVA of H index fit1 < - aov(H~recycle, data = alphadiversity) summary(fit1) #Provide P - value Tukey1 < - TukeyHSD(fit1, conf.level=0.95) #Tukey multiple comparison Tukey1 #Plot Tukey results #ANOVA of D index fit2 < - aov(D~recycle, data = alphadiversity) summary(fit2) #Provide P - value Tukey2 < - TukeyHSD(fit2, conf.level=0.95) #Tukey multiple comparison Tukey2 #Plot Tukey results #ANOVA of iD index fit3 < - aov(iD~recycle, data = alphadiversity) summary(fit3) #Provide P - value Tukey3 < - TukeyHSD(fit3, conf.level=0.95) #Tukey multiple comp arison Tukey3 #Plot Tukey results #ANOVA of J index fit4 < - aov(J~recycle, data = alphadiversity) summary(fit4) #Provide P - value Tukey4 < - TukeyHSD(fit4, conf.level=0.95) #Tukey multiple comparison Tukey4 #Plot Tukey results ##boxplot of H and J and D an d iD box_1 < - ggboxplot(alphadiversity, x = "recycle", y = "H", color="recycle")+ ylab("Shannon's Index (H)") + ylim(0, 1)+ theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 11, family= "Times New Roman"), axis.text.x = element_text(size = 14, family="Times New Roman", angle = 0, hjust = 0.5), axis.title.y = element_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, f amily="Times New Roman"), legend.title= element_blank(), legend.direction="vertical") box_2 < - ggboxplot(alphadiversity, x = "recycle", y = "J", color="recycle")+ ylab("Pielou's Index (J)") + ylim(0, 0.5) + theme( legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 11, family="Times New Roman"), axis.text.x = element_text(size = 14, family="Times New Roman", angle = 0, hjust = 0.5), axis.title.y = elemen t_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_blank(), legend.direction="vertical") box_3 < - ggboxplot(alphadiversity, x = "recycle", y = "D", c olor="recycle")+ ylab("Simpson (D)") + ylim(0, 0.5) + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 11, family="Times New Roman"), axis.text.x = element_text(size = 14, family="Times New Roman", angle = 0, hjust = 0.5), axis.title.y = element_text(size = 14, family="Times New Roman"), 76 legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_blank(), legend.direction="vertical") box_4 < - ggboxplot(alphadiversity, x = "recycle", y = "iD", color="recycle")+ ylab("Inverse Simpson (iD)") + ylim(0, 2) + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 11, family="Times New Rom an"), axis.text.x = element_text(size = 14, family="Times New Roman", angle = 0, hjust = 0.5), axis.title.y = element_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_blank(), legend.direction="vertical") grid.arrange(box_1, box_2, box_3, box_4, nrow=2) #Rarefaction ----------------------------------------------- col < - c("black", "darkred", "forestgreen", "orange", "blue", "yellow", "hotpink") lty < - c("solid", "dashed", "longdash", "dotdash") pars < - expand.grid(col = col, lty = lty, stringsAsFactors = FALSE) ra < - rarecurve(t.Frequency.table, step = 20, col =col,lty = lty, cex = 0.6) # Rarefaction Curve rad < - rad.lognorma l(t.Frequency.table) # Rank of Abundance rad1 < - plot(rad, xlab = "Rank", ylab = "Abundance") # Plotting the rank # Beta diversity --------------------------------------------------------- # Dend r ogram ----------------------------------------------------- ---------- par(mfrow=c(1,1)) distance < - vegdist(t.Frequency.table, method="euclidean") ## Generate d istance m atrix cluster < - hclust(distance, method="complete", members = NULL) ## Production of Hierarchical Cluster Production tree_m < - plot(cluster, xlab = "Samples", sub = NULL, main ="Dend r ogram") range(distance) rect.hclust(cluster, k = 3, border = "red") grp < - cutree(cluster, k = 3) P art B ( recirculation dynamics ), statistics ## Metagenomic analysis ## Algal Cultivation: Dynamic analysis of Recirculat ion vs. Freshwater ## Part B ## Wei Liao, January, 2020 update ## Carly Daiek, February 2020 update # Install "phyloseq" package # source ('http://bioconductor.org/biocLite.R') # biocLite('phyloseq') # Loading Library and Tables ---------------------------------------------- library(vegan) library (MASS) library(ggplot2) library(grid) library(gridExtra) library(ggpubr) library(plyr) library(RVAideMemoire) 77 library(DescTools) library(PMCMRplus) library(tadaatoolbox) library(inferr ) #+++++++++++++++++++++++++ # Function to calculate the mean and the standard deviation # for each group #+++++++++++++++++++++++++ # data : a data frame # varname : the name of a column containing the variable #to be summarized # groupnames : vector of column names to be used as # grouping variables data_summary < - function(data, varname, groupnames){ require(plyr) summary_func < - function(x, col){ c(mean = mean(x[[col]], na.rm=TRUE), sd = sd(x[[col]], na.rm=TRUE)) } data_sum< - ddply(data, groupnames, .fun=summary_func, varname) data_sum < - rename(data_sum, c("mean" = varname)) return(data_sum) } ## the .txt file needs to be saved as the type of "Tab delimited". ##Choose the Relative Frequency Table shoul d be a .txt con < - file.choose(new = FALSE) ##Now choose the Frequency Table Taxanomy should be .txt con1 < - file.choose(new = FALSE) ##Now choose the Meta data table should be .txt con2 < - file.choose(new = FALSE) Frequency_Table < - read.table(con, header = T, row.names = 1) Frequency_Table_taxonomy < - read.delim(con1, header = T, row.names = 1) metadata < - read.table(con2, header = T, row.names = 1) #this table includes key OTU from the .csv files generated in the following analysis metadata$month < - fact or(metadata$month) metadata$recycle < - factor(metadata$recycle) ## Abundances ---------------------------------------------------- #Phyloseq Full_Frequency < - cbind.data.frame(Frequency_Table, Frequency_Table_taxonomy) Frequency < - otu_table(Frequency_Ta ble,taxa_are_rows = TRUE) # Frequency Table production for phyloseq TAX < - tax_table(as.matrix(Frequency_Table_taxonomy)) ## Taxanomy production for phyloseq physeq < - phyloseq(Frequency, TAX) ##physeq document production physeq0 < - tax_glom(physeq, taxrank=rank_names(physeq)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) tax_table(physeq0) ## Overall abundances for Domain, Phylum, Class, Order, and Family --------- # Abundance Plotbar Domain physeqa < - tax_glom(physeq, taxrank=rank_names(physeq)[1], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) 78 tablea < - otu_table(physeqa) write.csv(tablea, "domain.csv") #Abundance Plotbar Phylum physeqa1 < - tax_glom(physeq, taxrank=rank_names(physeq)[2], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) tablea1 < - otu_table(p hyseqa1) write.csv(tablea1, "Phylum.csv") #Abundance Plotbar Class physeqa2 < - tax_glom(physeq, taxrank=rank_names(physeq)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) tablea2 < - otu_table(physeqa2) write.csv(tablea2, "Class.csv") ## Abundance Plotbar B acteria ------------ #Abundance Plotbar Bacteria (Phylum) physeq2 < - subset_taxa(physeq, Domain== "Bacteria") physeq2_1 < - tax_glom(physeq2, taxrank=rank_names(physeq2)[2], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) table2_1 < - otu_table(physeq2_1) write.csv(table2_1, "bacterialPhylum.csv") ##Abundance Plotbar Bacteroidetes (Class) physeq3 < - subset_taxa(physeq, Phylum == "Bacteroidetes") physeq3_1 < - tax_glom(physeq3, taxrank=rank_names(physeq3)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) table3_1 < - otu_table(physeq3_1) write.csv(table3_1, "BacteroidetesFamily.csv") #Abundance Plotbar Proteobacteria (Class) physeq5 < - subset_taxa(physeq, Phylum == "Proteobacteria") physeq5_1 < - tax_glom(physeq5, taxrank=rank_names(physeq5)[3], NArm=TRUE, bad_empty= c(NA, "", " ", " \ t")) table5_1 < - otu_table(physeq5_1) write.csv(table5_1, "ProteobacteriaFamily.csv") # ANOVA Eukarya (Domain) Eukarya < - aov(Domain.Eukarya~month, data = metadata) summary(Eukarya) data_summary(metadata, varname="Domain.Eukarya", groupnames=c("month")) # ANOVA Eukarya (Domain) Bacteria < - aov(Domain.Bacteria~month, data = metadata) summary(Bacteria) data_summary(metadata, varname="Domain.Bacteria", groupnames=c("month")) # ANOVA Proteobacteria (Phylum) Proteobacteria < - aov(Phylum.Proteobacteria~month, data = metadata) summary(Proteobacteria) data_summary(metadata, varname="Phylum.Proteobacteria", groupnames=c("month")) TukeyHSD(Proteobacteria) #If ANOVA reports P < 0.05, use TukeyHSD to detect differences between monthly comparisons 79 #ANOVA Bacteroidetes (Phylum) Bacteroidetes < - aov(Phylum.Bacteroidetes~month, data = metadata) summary(Bacteroidetes) data_summary(metadata, varname="Phylum.Bacteroidetes", groupnames=c("month")) TukeyHSD(Bacteroidetes) #ANOVA BacteriaUnclassified (Phylum) BacteriaUnclassified < - aov(Phylum.BacteriaUnclassified~month, data = metadata) summary(BacteriaUnclassified) data_summary(metadata, varname="Phylum.BacteriaUnclass ified", groupnames=c("month")) TukeyHSD(BacteriaUnclassified) # ANOVA Alphaproteobacteria (Class) Alphaproteobacteria < - aov(Class.Alphaproteobacteria~month, data = metadata) summary(Alphaproteobacteria) data_summary(metadata, varname="Class.Alphaproteobacteria", groupnames=c("month")) # ANOVA Betaproteobacteria (Class) Betaproteobacteria < - aov(Class.Betaproteobacteria~month, data = metadata) summary(Betaproteobacteria) data_summary(metadata, varname="Class.Beta proteobacteria", groupnames=c("month")) data_summary(metadata, varname="Class.Betaproteobacteria", groupnames=c("recycle")) TukeyHSD(Betaproteobacteria) # ANOVA Gammaproteobacteria (Class) Gammaproteobacteria < - aov(Class.Gam maproteobacteria~month, data = metadata) summary(Gammaproteobacteria) data_summary(metadata, varname="Class.Gammaproteobacteria", groupnames=c("month")) # ANOVA Proteobacteria_unclassified (Class) ProteobacteriaUnclassified < - aov(Class.Pro teobacteriaUnclassified~month, data = metadata) summary(ProteobacteriaUnclassified) data_summary(metadata, varname="Class.ProteobacteriaUnclassified", groupnames=c("month")) TukeyHSD(ProteobacteriaUnclassified) # ANOVA Bacteroidetes_unclassified (Class) BacteroidetesUnclassified < - aov(Class.BacteroidetesUnclassified~month, data = metadata) 80 summary(BacteroidetesUnclassified) data_summary(metadata, varname="Class.BacteroidetesUnclassified", groupnames=c("mo nth")) TukeyHSD(BacteroidetesUnclassified) # ANOVA [Saprospirae] (Class) Saprospirae < - aov(Class.Saprospirae~month, data = metadata) summary(Saprospirae) data_summary(metadata, varname="Class.Saprospirae", groupnames=c("month")) TukeyHSD (Saprospirae) # ANOVA Cytophagia (Class) Cytophagia < - aov(Class.Cytophagia~month, data = metadata) summary(Cytophagia) data_summary(metadata, varname="Class.Cytophagia", groupnames=c("month")) TukeyHSD(Cytophagia) # ANOVA Flavobacteria (Class) Flavobacteria < - aov(Class.Flavobacteria~month, data = metadata) summary(Flavobacteria) data_summary(metadata, varname="Class.Flavobacteria", groupnames=c("month")) TukeyHSD(Flavobacteria) # ANOVA Sphingobacteria (Cla ss) Sphingobacteria < - aov(Class.Sphingobacteria~month, data = metadata) summary(Sphingobacteria) data_summary(metadata, varname="Class.Sphingobacteria", groupnames=c("month")) TukeyHSD(Sphingobacteria) P art B ( recirculation dynamics ), plotting ## Metagenomic analysis ## Algal Cultivation: Dynamic analysis of Recirculation vs. Freshwater ## Part B - PLOTTING ## Wei Liao, Janurary, 2020 update ## Carly Daiek, February 2020 update # Install "phyloseq" package # source ('http://bioconductor.org/biocLite.R') # biocLite('phyloseq') 81 # Loading Library and Tables ---------------------------------------------- library(vegan) library(phyloseq) library (MASS) library(ggplot2) library(grid) library(gridExtra) library(ggpubr) # Installing the font package --------------------------------------------- library(extrafont) font_import() #It may take a few minutes to import. loadfonts(device="win") ## the .txt file needs to be saved as the type of "Tab delimited". ##Choose the Relative Frequency Table should be a .txt con < - file.choose(new = FALSE) ##Now choose the Frequency Table Taxanomy should be .txt con1 < - file.choose(new = FALSE) ##Now choose the Meta data table should be .txt ##con2 < - file.choose(new = FALSE) metadata < - read.table(con, header = T, row.names = 1) Frequency_Table < - metadata order < - c("Month 1", "Month 2", "Month 3", "Month 4", "Freshwater") order < - factor(order,levels = c("Month 1", "Month 2", "Month 3", "Month 4", "Freshwater")) names(Freque ncy_Table) < - order Frequency_Table_taxonomy < - read.delim(con1, header = T, row.names = 1) ## Abundances ---------------------------------------------------- #Phyloseq Full_Frequency < - cbind.data.frame(Frequency_Table, Frequency_Table_taxonomy) Frequen cy < - otu_table(Frequency_Table,taxa_are_rows = TRUE) # Frequency Table production for phyloseq TAX < - tax_table(as.matrix(Frequency_Table_taxonomy)) ## Taxanomy production for phyloseq #SAM < - sample_data(metadata) physeq < - phyloseq(Frequency, TAX) ##phy seq document production physeq0 < - tax_glom(physeq, taxrank=rank_names(physeq)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) tax_table(physeq0) p = plot_bar(physeq0, fill = "Class", facet_grid=Domain~Phylum, ) + theme(axis.title.x = element_blank(), ax is.text.x = element_text(size = 5, angle = 45, hjust = 1)) + geom_bar(color = "black", size = .1, stat = "identity", position = "stack") p ## Overall abundances for Domain, Phylum, Class, Order, and Family --------- # Abundance Plotbar Domain physeqa < - tax_glom(physeq, taxrank=rank_names(physeq)[1], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) tablea < - otu_table(physeqa) tablea write.csv(tablea, "domain.csv") 82 a = plot_bar(physeqa, fill = "Domain") + geom_bar(aes(color=Domain, fill=Domain), st at = "identity", position = "stack") + ylab("Relative Frequency (%)") + labs(title= "") + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 12, family="Times New Roman"), axis.text.x = ele ment_text(size = 10, family="Times New Roman", angle = 45, hjust = 1), axis.title.y = element_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_text(siz e = 12, family="Times New Roman"), legend.direction="vertical") a #Abundance Plotbar Phylum physeqa1 < - tax_glom(physeq, taxrank=rank_names(physeq)[2], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) tablea1 < - otu_table(physeqa1) #tablea1 write.csv(ta blea1, "Phylum.csv") a1 = plot_bar(physeqa1, fill = "Phylum") + geom_bar(aes(color=Phylum, fill=Phylum), stat = "identity", position = "stack") + ylab("Relative Frequency (%)") + labs(title= "") + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 12, family="Times New Roman"), axis.text.x = element_text(size = 10, family="Times New Roman", angle = 45, hjust = 1), axis.title.y = element_text(size = 14, family=" Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_text(size = 12, family="Times New Roman"), legend.direction="vertical") a1 #Abundance Plotbar Class physeqa2 < - tax_glom(phys eq, taxrank=rank_names(physeq)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) tablea2 < - otu_table(physeqa2) tablea2 write.csv(tablea2, "Class.csv") a2 = plot_bar(physeqa2, fill = "Class") + geom_bar(aes(color=Class, fill=Class), stat = "identity", posi tion = "stack") + ylab("Relative Frequency (%)") + labs(title= "") + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 12, family="Times New Roman"), axis.text.x = element_text(size = 10, family="Times New Roman", angle = 45, hjust = 1), axis.title.y = element_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_text(size = 12, family="Times New Roman"), legend.direction="vertical") a2 grid.arrange(a,a1,a2,nrow=1) ## Abundance Plotbar Bacteria ------------ #Abundance Plotbar Bacteria (Phylum) physeq2 < - subset_taxa(physeq, Domain== "Bacteria") physeq2_1 < - tax_glom (physeq2, taxrank=rank_names(physeq2)[2], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) 83 table2_1 < - otu_table(physeq2_1) table2_1 write.csv(table2_1, "bacterialPhylum.csv") c = plot_bar(physeq2_1, fill = "Phylum") + geom_bar(aes(color=Phylum, fill=Phylum), s tat = "identity",position = "stack") + ylab("Relative Frequency (%)") + labs(title = "") + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 12, family="Times New Roman"), axis.text.x = ele ment_text(size = 10, family="Times New Roman", angle = 45, hjust = 1), axis.title.y = element_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_text(siz e = 12, family="Times New Roman"), legend.direction="vertical") c ##Abundance Plotbar Bacteroidetes (Class) physeq3 < - subset_taxa(physeq, Phylum == "Bacteroidetes") physeq3_1 < - tax_glom(physeq3, taxrank=rank_names(physeq3)[3], NArm=TRUE, bad_empt y=c(NA, "", " ", " \ t")) table3_1 < - otu_table(physeq3_1) table3_1 write.csv(table3_1, "BacteroidetesFamily.csv") d = plot_bar(physeq3_1, fill = "Class")+ geom_bar(aes(color=Class, fill=Class), stat = "identity",position = "stack") + ylab ("Bacteroidetes Abundance (%)") + xlab("Samples") + labs(title = "") + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 12, family="Times New Roman"), axis.text.x = element_text(size = 10, f amily="Times New Roman", angle = 45, hjust = 1), axis.title.y = element_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_text(size = 12, family="Times New Roman"), legend.direction="vertical") d #Abundance Plotbar Proteobacteria (Class) physeq5 < - subset_taxa(physeq, Phylum == "Proteobacteria") physeq5_1 < - tax_glom(physeq5, taxrank=rank_names(physeq5)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t" )) table5_1 < - otu_table(physeq5_1) table5_1 write.csv(table5_1, "ProteobacteriaFamily.csv") f = plot_bar(physeq5_1, fill = "Class")+ geom_bar(aes(color=Class, fill=Class), stat = "identity",position = "stack") + ylab("Proteobacteria Abundance (%)") + xlab("Samples") + labs(title = "") + theme(legend.position="right", axis.title.x = element_blank(), axis.text.y = element_text(size = 12, family="Times New Roman"), axis.text.x = element_text(size = 10, family="Times New Roman", angle = 4 5, hjust = 1), axis.title.y = element_text(size = 14, family="Times New Roman"), legend.text = element_text(size = 11, family="Times New Roman"), legend.title= element_text(size = 12, family="Times New Roman"), legend.direc tion="vertical") f grid.arrange(d,f,nrow=1) 84 ## Heatmap --------------------- heatorder < - order #Heatmap Phylum in bacteria physeq9 < - subset_taxa(physeq, Domain== "Bacteria") physeq9_1 < - tax_glom(physeq9, taxrank=rank_names(physeq9)[2], NArm =TRUE, bad_empty=c(NA, "", " ", " \ t")) i = plot_heatmap(physeq9_1, method = "NMDS", distance = "bray", sample.label = NULL, taxa.label = "Phylum", low = "#00cd00", high = "#003400", na.value = "white", max.label = 250, title = NULL, sample.order = heatorder, taxa.order = NULL, first.sample = NULL, first.taxa = NULL)+ theme(legend.position="right", axis.title.x = element_blank(), axis.text.x = element_text(size = 10, angle = 45, hjust = 0.70), axis.title.y = element_text(size = 12), axis.text.y = element_text(size = 10), legend.text = element_text(size = 10), legend.title= element_text(size = 12), plot.title= element_text(size = 15)) i #Heatmap family in bacteroidetes physeq10 < - subset_taxa(physeq, Phylum == "Bacteroidetes") physeq10_1 < - tax_glom(physeq10, taxrank=rank_names(physeq10)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) j = plot_heatmap(physeq10_1, method = "NMDS", distan ce = "bray", sample.label = NULL, taxa.label = "Class", low = "#FFCCCB", high = "#8B0000", na.value = "white", max.label = 250, title = NULL, sample.order = heatorder, taxa.order = NULL, first.sample = NULL, first.taxa = NULL) + theme(legend.position="right", axis.title.x = element_blank(), axis.text.x = element_text(size = 10, angle = 45, hjust = 0.70), axis.title.y = element _text(size = 12), axis.text.y = element_text(size = 10), legend.text = element_text(size = 10), legend.title= element_text(size = 12), plot.title= element_text(size = 15)) j #Heatmap Proteobacteria physeq11 < - subset_taxa(physeq, Phylum== "Proteobacteria") physeq11_1 < - tax_glom(physeq11, taxrank=rank_names(physeq11)[3], NArm=TRUE, bad_empty=c(NA, "", " ", " \ t")) k = plot_heatmap(physeq11_1, method = "NMDS", distance = "bray", sample.label = N ULL, taxa.label = "Class", low = "#66CCFF", high = "#000033", na.value = "white", max.label = 250, title = NULL, sample.order = heatorder, taxa.order = NULL, first.sample = NULL, first.taxa = NULL)+ t heme(legend.position="right", axis.title.x = element_blank(), axis.text.x = element_text(size = 10, angle = 45, hjust = 0.70), axis.title.y = element_text(size = 12), axis.text.y = element_text(size = 10), legend.text = element_text(size = 10), legend.title= element_text(size = 12), 85 plot.title= element_text(size = 15)) k grid.arrange(i,j,k,nrow=1) 86 REFERENCES 87 REFERENCES 1. Agency, I.E. (2019) Global Energy & CO2 Status Report 2018. In: The latest trends in energy and emissions in 2018: IEA Publications. 2. Cheah, W.Y., Show, P.L., Chang, J.S., Ling, T.C., and Juan, J.C. (2015) Biosequestration of atmospheric CO2 and flue gas - containing CO2 by microalgae. Bioresource Technology 184 , 190 - 201. 3. Singh, J., and Dhar, D.W. (2019) Overview of Carbon Capture Technology: Microalgal Biorefinery Concept and State - of - the - Art. Frontiers in Marine Science 6 . 4. Khan, M.I., Shin, J.H., and Kim, J.D. (2018) The promising future of microalgae: current status, challenges, and optimization of a sustainable and renewable industry for biofuels, feed, and other products. Microbial Cell Factories 17 . 5. Bhola, V., Swalaha, F., Kumar, R.R., Singh, M., and Bux, F. (2014) Overview of the potential of microalgae for CO 2 sequestration. International Journal of Environmental Science and Technology 11 , 2103 - 2118. 6. Abd Rahaman, M.S., Cheng, L. - H., Xu, X. - H., Zhang, L., and Chen, H. - L. (2011) A review of carbon dioxide capture and utilization by membrane integrated microa lgal cultivation processes. Renewable & Sustainable Energy Reviews 15 , 4002 - 4012. 7. Narala, R.R., Garg, S., Sharma, K.K., Thomas - Hall, S.R., Deme, M., Li, Y., and Schenk, P.M. (2016) Comparison of Microalgae Cultivation in Photobioreactor, Open Raceway P ond, and a Two - Stage Hybrid System. Frontiers in Energy Research 4 . 8. Fulbright, S.P., Robbins - Pianka, A., Berg - Lyons, D., Knight, R., Reardon, K.F., and Chisholm, S.T. (2018) Bacterial community changes in an industrial algae production system. Algal Re search - Biomass Biofuels and Bioproducts 31 , 147 - 156. 9. Cruz, Y.R., Aranda, D.A.G., Seidl, P.R., Diaz, G.C., Carliz, R.G., Fortes, M.M., da Ponte, D.A.M.P., and de Paula , R.C.V. (2018) Cultivation Systems of Microalgae for the Production of Biofuels. 1 0. Rosenberg, J.N., Mathias, A., Korth, K., Betenbaugh, M.J., and Oyler, G.A. (2011) Microalgal biomass production and carbon dioxide sequestration from an integrated ethanol biorefinery in Iowa: A technical appraisal and economic feasibility evaluation. B iomass & Bioenergy 35 , 3865 - 3876. 11. Slade, R., and Bauen, A. (2013) Micro - algae cultivation for biofuels: Cost, energy balance, environmental impacts and future prospects. Biomass & Bioenergy 53 , 29 - 38. 88 12. Silva, P.P.d., and Ribeiro, L.A. (2019) Chapt er 19 - Assessing Microalgae Sustainability as a Feedstock for Biofuels. Woodhead Publishing Series in Energy, 373 - 392. 13. Shahid, A., Sana, M., Asraful, A.M., Nazia, N., and Muhammad, M. (2019) The Culture Technology for Freshwater and Marine Microalg ae. 21 - 44. 14. Fret, J., Roef, L., Diels, L., Tavernier, S., Vyverman, W., and Michiels, M. (2020) Combining medium recirculation with alternating the microalga production strain: a laboratory and pilot scale cultivation test. Algal Research - Biomass Biofu els and Bioproducts 46 . 15. Lian, J., Wijffels, R.H., Smidt, H., and Sipkema, D. (2018) The effect of the algal microbiome on industrial production of microalgae. Microbial Biotechnology 11 , 806 - 818. 16. Padmaperuma, G., Kapoore, R.V., Gilmour, D.J., and Vaidyanathan, S. (2018) Microbial consortia: a critical look at microalgae co - cultures for enhanced biomanufacturing. Critical Reviews in Biotechnology 38 , 690 - 703. 17. Luis Fuentes, J., Garbayo, I., Cuaresma, M., Montero, Z., Gonzalez - del - Valle, M., and Vilchez, C. (2016) Impact of Microalgae - Bacteria Interactions on the Production of Algal Biomass and Associated Compounds. Marine Drugs 14 . 18. Yao, S., Lyu, S., An, Y., Lu, J., Gjermansen, C., and Schramm, A. (2019) Microalgae - bacteria symbiosis in microalgal growth and biofuel production: a review. Journal of Applied Microbiology 126 , 359 - 368. 19. Soman, A., and Shastri, Y. (2015) Optimization of novel photobioreactor design using computational fluid dynamics. Applied Energy 140 , 246 - 255. 20. Zheng, Y.B., Li, T.T., Yu, X.C., Bates, P.D., Dong, T., and Chen, S.L. (2013) High - density fed - batch cultur e of a thermotolerant microalga Chlorella sorokiniana for biofuel production. Applied Energy 108 , 281 - 287. 21. Morita, M., Watanabe, Y., and Saiki, H. (2000) High photosynthetic productivity of green microalga Chlorella sorokiniana. Applied Biochemistry a nd Biotechnology 87 , 203 - 218. 22. Sorokin, C., and Myers, J. (1953) A HIGH - TEMPERATURE STRAIN OF CHLORELLA. Science 117 , 330 - 331. 23. Carfagna, S., Lanza, N., Salbitani, G., Basile, A., Sorbo, S., and Vona, V. (2013) Physiological and morphological respo nses of Lead or Cadmium exposed Chlorella sorokiniana 211 - 8K (Chlorophyceae). Springerplus 2 . 24. Lizzul, A.M., Lekuona - Amundarain, A., Purton, S., and Campos, L.C. (2018) Characterization of Chlorella sorokiniana, UTEX 1230. Biology - Basel 7 . 89 25. de - Bash an, L.E., Trejo, A., Huss, V.A.R., Hernandez, J. - P., and Bashan, Y. (2008) Chlorella sorokiniana UTEX 2805, a heat and intense, sunlight - tolerant microalga with potential for removing ammonium from wastewater. Bioresource Technology 99 , 4980 - 4989. 26. Liz zul, A.M., Lekuona - Amundarain, A., Purton, S., and Campos, L.C. (2018) Characterization of Chlorella sorokiniana, UTEX 1230. Biology - Basel 7 , 25 - Article No.: 25. 27. Zhang, B., Li, W., Guo, Y., Zhang, Z.Q., Shi, W.X., Cui, F.Y., Lens, P.N.L., and Tay, J.H . (2020) Microalgal - bacterial consortia: From interspecies interactions to biotechnological applications. Renewable & Sustainable Energy Reviews 118 . 28. Amavizca, E., Bashan, Y., Ryu, C. - M., Farag, M.A., Bebout, B.M., and de - Bashan, L.E. (2017) Enhanced performance of the microalga Chlorella sorokiniana remotely induced by the plant growth - promoting bacteria Azospirillum brasilense and Bacillus pumilus. Scientific reports 7 , 41310. 29. Rodolfi, L., Zittelli, G.C., Barsanti, L., Rosati, G., and Tredici, M .R. (2003) Growth medium recycling in Nannochloropsis sp mass cultivation. Biomolecular Engineering 20 , 243 - 248. 30. Wan, M.X., Jin, X.J., Xia, J.L., Rosenberg, J.N., Yu, G., Nie, Z.Y., Oyler, G.A., and Betenbaugh, M.J. (2014) The effect of iron on growth , lipid accumulation, and gene expression profile of the freshwater microalga Chlorella sorokiniana. Applied Microbiology and Biotechnology 98 , 9473 - 9481. 31. Chimiklis, P.E., and Karlander, E.P. (1973) LIGHT AND CALCIUM INTERACTIONS IN CHLORELLA INHIBITE D BY SODIUM - CHLORIDE. Plant Physiology 51 , 48 - 56. 32. Tousova, Z., Froment, J., Oswald, P., Slobodnik, J., Hilscherova, K., Thomas, K.V., Tollefsen, K.E., Reid, M., Langford, K., and Blaha, L. (2018) Identification of algal growth inhibitors in treated waste water using effect - directed analysis based on non - t arget screening techniques. Journal of Hazardous Materials 358 , 494 - 502. 33. Liang, S., Kang, Y., Zeng, L., Qin, Y., Wang, L., Zhang, Q., and Luo, J. (2017) How Chlorella sorokiniana and its High Tolerance to Pb Might be a Potential Pb Biosorbent. Polish Journal of Environmental Studies 26 , 1139 - 1146. 34. Spence, W.H. (2016) REGROWTH OF CHLORELLA SOROKINIANA ON RECYCLED MEDIA WITH REPLENISHED NUTRIENTS. 35. Zhou, W.L., Qiao, X.T., Sun, J.F., Xing, K.Z., and Tang, X.X. (2011) Ecological Effect of Z - QS01 Strain on Chlorella vulgaris and Its Response to UV - B Radiation Stress. 2011 2nd International Conference on Challenges in Environmental Science and Computer E ngineering (Cesce 2011), Vol 11, Pt B 11 , 741 - 748. 90 36. Ikawa, M., Sasner, J.J., and Haney, J.F. (1997) Inhibition of Chlorella growth by degradation and related products of linoleic and linolenic acids and the possible significance of polyunsaturated fatt y acids in phytoplankton ecology. Hydrobiologia 356 , 143 - 148. 37. Richmond, A. (2003) Growth characteristics of ultrahigh - density microalgal cultures. Biotechnology and Bioprocess Engineering 8 , 349 - 353. 38. Richmond, A. (2004) Handbook of Microalgal Cul ture: Biotechnology and Applied Phycology. Blackwell Science. 39. Pratt, R., and Fong, J. (1940) Studies on Chlorella vulgaris II Further evidence that Chlorella cells form a growth - inhibiting substance. American Journal of Botany 27 , 431 - 436. 40. Casper s, H. (1966) Fogg, G. E.: Algal Cultures and Phytoplankton Ecology. 1965. With 31 Fig., 3 pl. Madison and Milwaukee: University of Wisconsin Press 1965. 1266 pp. Internationale Revue der gesamten Hydrobiologie und Hydrographie 51 , 203 - 204. 41. Spoehr, H .A., Smith, J.H.C., Strain, H.H., Milner, H.W., and Hardin, G.J. (1949) Fatty acid antibacterials from plants. Carnegie Inst Washington Publ 586 , 1 - 67. 42. Rishiram Ramanan , B. - H.K., Dae - Hyun Cho , Hee - Mock Oh , Hee - Sik Kim. (January - Februa ry 2016) Algae bacteria interactions: Evolution, ecology and emerging applications. 34 1 , 14 - 29. 43. Mandalam, R.K., and Palsson, B.O. (1995) CHLORELLA - VULGARIS (CHLORELLACEAE) DOES NOT SECRETE AUTOINHIBITORS AT HIGH CELL DENSITIES. American Journal of Bo tany 82 , 955 - 963. 44. Sakata, T., Yoshikawa, T., and Nishitarumizu, S. (2011) Algicidal activity and identification of an algicidal substance produced by marine Pseudomonas sp. C55a - 2. Fisheries Science 77 , 397 - 402. 45. Nakashima, T., Miyazaki, Y., Matsu yama, Y., Muraoka, W., Yamaguchi, K., and Oda, T. (2006) Producing mechanism of an algicidal compound against red tide phytoplankton in a - proteobacterium. Applied Microbiology and Biotechnology 73 , 684 - 690. 46. Routledge, S.J. (2012) BE YOND DE - FOAMING: THE EFFECTS OF ANTIFOAMS ON BIOPROCESS PRODUCTIVITY. Computational and Structural Biotechnology Journal 3 . 47. Al - Masry, W.A. (1999) Effects of antifoam and scale - up on operation of bioreactors. Chemical Engineering and Processing - Process Intensification 38 , 197 - 201. 48. Koch, V., Ruffer, H.M., Schugerl, K., Innertsberger, E., Menzel, H., and Weis, J. (1995) EFFEC T OF ANTIFOAM AGENTS ON THE MEDIUM AND MICROBIAL CELL PROPERTIES AND PROCESS PERFORMANCE IN SMALL AND LARGE REACTORS. Process Biochemistry 30 , 435 - 446. 91 49. Wang, X.Q., Ruan, Z.H., Sheridan, P., Boileau, D., Liu, Y., and Liao, W. (2015) Two - stage photoautotrophic cultivation to improve carbohydrate production in Chlamydomonas reinhardtii. Biomass & Bioenergy 74 , 280 - 287. 50. Cutshaw, A., Daiek, C . , Zheng, Y., Frost, H., Marks, A., Clements, D., Uludag - Demirer, S., Verhanovitz, N., Pavlik, D., Clary, W., Liu, Y., and Liao, W. (2020) A long - term pilot - scale algal cultivation on power plant flue gas Cultivation stability and biomass accumulation [under developm ent]. 51. Pavlik, D., Zhong, Y., Daiek, C., Liao, W., Morgan, R., Clary, W., and Liu, Y. (2017) Microalgae cultivation for carbon dioxide sequestration and protein production using a high - efficiency photobioreactor system. Algal Research - Biomass Biofuels and Bioproducts 25 , 413 - 420. 52. Company, H. (1996) EPA Compliant Methods 53. Bolyen, E., Rideout, J.R., Dillon, M.R., Bokulich, N., Abnet, C.C., Al - Ghalith, G.A., Alexander, H., Alm, E.J., Arumugam, M., Asnicar, F., Bai, Y., Bisanz, J.E., Bittinger, K. , Brejnrod, A., Brislawn, C.J., Brown, C.T., Callahan, B.J., Caraballo - Rodriguez, A.M., Chase, J., Cope, E.K., Da Silva, R., Diener, C., Dorrestein, P.C., Douglas, G.M., Durall, D.M., Duvallet, C., Edwardson, C.F., Ernst, M., Estaki, M., Fouquier, J., Gaug litz, J.M., Gibbons, S.M., Gibson, D.L., Gonzalez, A., Gorlick, K., Guo, J.R., Hillmann, B., Holmes, S., Holste, H., Huttenhower, C., Huttley, G.A., Janssen, S., Jarmusch, A.K., Jiang, L.J., Kaehler, B.D., Bin Kang, K., Keefe, C.R., Keim, P., Kelley, S.T., Knights, D., Koester, I., Kosciolek, T., Kreps, J., Langille, M.G.I., Lee, J., Ley, R., Liu, Y.X., Loftfield, E., Lozupone, C., Maher, M., Marotz, C., Martin, B.D., McDonald, D., McIver, L.J., Melnik, A.V., Metcalf, J.L., Morgan, S.C., Morton, J.T., Naime y, A.T., Navas - Molina, J.A., Nothias, L.F., Orchanian, S.B., Pearson, T., Peoples, S.L., Petras, D., Preuss, M.L., Pruesse, E., Rasmussen, L.B., Rivers, A., Robeson, M.S., Rosenthal, P., Segata, N., Shaffer, M., Shiffer, A., Sinha, R., Song, S.J., Spear, J .R., Swafford, A.D., Thompson, L.R., Torres, P.J., Trinh, P., Tripathi, A., Turnbaugh, P.J., Ul - Hasan, S., vander Hooft, J.J.J., Vargas, F., Vazquez - Baeza, Y., Vogtmann, E., von Hippel, M., Walters, W., Wan, Y.H., Wang, M.X., Warren, J., Weber, K.C., Willi amson, C.H.D., Willis, A.D., Xu, Z.Z., Zaneveld, J.R., Zhang, Y.L., Zhu, Q.Y., Knight, R., and Caporaso, J.G. (2019) Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nature Biotechnology 37 , 852 - 857. 54. Sager, J.C ., and McFarlane, J.C. (1997) Radiation. Iowa State University: North Central Regional Research Publication No. 340, Iowa Agriculture and Home Economics Experiment Station Special Report no. 99. 55. Ghannadzadeh, A. (2013) Exergetic balances and analysis in a Process Simulator: A way to enhance Process Energy Integration. In: Chemical Engineering Laboratory Toulouse, France: Institut National Polytechnique de Toulouse. 56. Burja, A.M., Tamagnini, P., Bustard, M.T., and Wright, P.C. (2001) Identification o f the green alga, Chlorella vulgaris (SDC1) using cyanobacteria derived 16S rDNA primers: targeting the chloroplast. Fems Microbiology Letters 202 , 195 - 203. 92 57. Sukenik, A., Zohary, T., and Padisák, J. (2009) Cyanoprokaryota and Other Prokaryotic Algae. In : Encyclopedia of Inland Waters. Likens, G.E. ed. Oxford: Academic Press, pp. 138 - 148. 58. Thomas, F., Hehemann, J. - H., Rebuffet, E., Czjzek, M., and Michel, G. (2011) Environmental and gut Bacteroidetes: the food connection. Frontiers in Microbiology 2 . 59. Rodrigues, R.R., Rodgers, N.C., Wu, X., and Williams, M.A. (2018) COREMIC: a web - tool to search for a niche associated CORE MICrobiome. Peerj 6 . 60. Huang, W., Li, B., Zhang, C., Zhang, Z., Lei, Z., Lu, B., and Zhou, B. (2015) Effect of algae growth on aerobic granulation and nutrients removal from synthetic wastewater by using sequencing batch reactors. Bioresource Technology 179 , 187 - 192. 61. Burrell, P.C., Phalen, C.M., and Hovanec, T.A. (2001) Identification of bacteria responsible for ammonia oxidation in freshwater aquaria. Applied and Environmental Microbiology 67 , 5791 - 5800. 62. Sambles, C., Moore, K., Lux, T.M., Jones, K., Littlejohn, G.R., Gouveia, J.D., Aves, S.J., Studholme, D.J., Lee, R., and Love, J. (2017) Metagenomic analysis of the complex microbial consortium associated with cultures of the oil - rich alga Botryococcus bra unii. Microbiologyopen 6 .