Trimethylamine abatement in algal-bacterial photobioreactors

1 Wordcount results: 6677 1 Trimethylamine abatement in algal-bacterial 2 photobioreactors 3 Celia Pascual1,2, Ilker Akmirza2,3, Rebeca Pérez1,2, Esther Arnaiz1,2, Raúl Muñoz1,2, 4 Raquel Lebrero1,2* 5 1Department of Chemical Engineering and Environmental Technology, University of Valladolid, Dr. 6 Mergelina s/n., Valladolid 47011, Spain 7 2Institute of Sustainable Processes, University of Valladolid, Dr. Mergelina, s/n, 47011, Valladolid, 8 Spain. 9 3Department of Environmental Engineering, Gebze Technical University, Kocaeli, 41400, Turkey 10 *-Author for correspondence: [email protected] 11 12 Keywords: Biodegradation, Bubble column bioreactor, Microalgae-bacteria, Odour 13 treatment; Photobioreactor, Trimethylamine. 14 15 Abstract 16 Trimethylamine (TMA) is an odorous volatile organic compound emitted by industries. Algal-17 based biotechnologies have been proven as a feasible alternative for wastewater treatment, 18 although their application to abate polluted air emissions is still scarce. This work 19 comparatively assessed the removal of TMA in a conventional bacterial bubble column 20 bioreactor (BC) and a novel algal-bacterial bubble column photobioreactor (PBC). The PBC 21 exhibited a superior TMA abatement performance compared to the conventional BC. In this 22 sense, the BC reached a removal efficiency (RE) and an elimination capacity (EC) of 78 % and 23 12.1 g TMA m-3 h-1, respectively, while the PBC achieved a RE of 97 % and a EC of 16.0 g 24 TMA m-3·h-1 at an empty bed residence time (EBRT) of 2 min and a TMA concentration ~500 25 mg m-3. The outstanding performance of the PBC allowed to reduce the operating EBRT to 1.5 26 and 1 min, while maintaining high REs of 98 and 94 %, and ECs of 21.2 and 28.1 g m-3·h-1, 27 respectively. Moreover, the PBC improved the quality of the gas and liquid effluents 28 discharged, showing a net CO2 consumption and decreasing by ~ 30 % the total nitrogen 29 concentration in the liquid effluent via biomass assimilation. A high specialization of the 30 bacterial community was observed in the PBC, Mumia and Aquamicrobium sp. being the most 31 abundant genus within the main phyla identified. 32 33 34 © Springer Nature Publishing AG. Environmental Science and Pollution Research volume 27, pages, 9028–9037 (2020) http:// dx.doi.org/10.1007/s11356-019-07369-z 2 35 1. Introduction 36 The widespread release of odorous emissions to the atmosphere has become crucial due 37 to their adverse effects on human health and the environment (Xue et al. 2013; Wei et 38 al. 2015). Among odorous compounds, trimethylamine (TMA, C3H9N) has been 39 identified as a potentially toxic and likely carcinogenic malodorous volatile organic 40 compound with a low odor threshold concentration of 0.2 µg m-3 (Chang et al., 2004). 41 Moreover, TMA exerts a detrimental effect on the synthesis of macromolecules such as 42 DNA, RNA and proteins (Liffourrena and Lucchesi 2014), besides inducing teratogenic 43 effects on animal embryos (Kim et al., 2003). TMA is emitted in wastewater treatment 44 and composting facilities, livestock farms and fish meal manufacturing plants; being 45 partially responsible for the unpleasant odor that characterizes these emissions (Chang 46 et al. 2004; Ding et al. 2008). A proper management of TMA-laden emissions according 47 to environmental regulatory limits is crucial not only to avoid safety and health hazards, 48 but also to mitigate environmental impacts (i.e. greenhouse effect, acid rain and 49 eutrophication) (Chang et al., 2004; Perillo and Rodríguez, 2016). 50 Biotechnologies have been consistently proven as cost-effective and environmentally 51 friendly alternatives to physical-chemical technologies for the abatement of odorous and 52 toxic gas pollutants (Ho et al. 2008; Estrada et al. 2011). Microorganisms belonging to 53 the genera Paracoccus, Hyphomicrobium, Pseudomonas, Methylophilus, Arthrobacter, 54 Aminobacter, Haloanaerobacter and Bacillus are capable of using TMA as the only 55 carbon and energy source (Ding et al. 2008). In addition, previous studies have 56 demonstrated the feasibility of biologically degrading TMA in packed bed bioreactors 57 such as biofilters and biotrickling filters, being the only biotechnology studied up to 58 date (Aguirre et al., 2018; Wan et al., 2011). However, even if high TMA removal rates 59 3 have been achieved in these systems, the accumulation of NH3 (end product of the 60 aerobic oxidation of TMA) typically induces the alkalization of the medium, which 61 ultimately limits TMA biodegradation (Ho et al. 2008). Previous researchers have also 62 evaluated the subsequent NH3 bio-oxidation to nitrite (NO2-) and nitrate (NO3-) by 63 heterotrophic bacteria (Oyarzun et al. 2019). However, these nitrogen-containing 64 species remain in the liquid phase resulting in a high nitrogen-loaded effluent. 65 In the past decades, algal-bacterial based technologies have been widely studied due to 66 their capacity to simultaneously degrading toxic and/or recalcitrant organic materials 67 and depleting nutrients such as ammonium and NO3at high removal rates (Borde et al. 68 2003; Muñoz and Guieysse 2006). In this context, processes based on the symbiotic 69 interaction between microalgae and bacteria may constitute a competitive alternative to 70 bacterial-based biotechnologies, where TMA is oxidized by bacteria and the N-NH3 71 released from TMA oxidation is fixed by microalgae. Microalgae also fix part of the 72 CO2 produced in the bacterial oxidation of TMA and provide oxygen during the 73 photosynthetic activity, thus reducing the CO2 footprint and the aeration needs of the 74 process (Kang et al. 2017). Moreover, the biomass generated in these processes can be 75 further valorized as biofuel feedstock or as biofertilizer (Muñoz and Guieysse, 2006). 76 Despite the above mentioned advantages, the implementation of algal-bacterial 77 processes for waste gas treatment has been scarcely studied. In this regard, the 78 configuration of the photobioreactor is of key relevance since it determines the 79 efficiency of light penetration in the algal-bacterial cultivation broth. Bubble column 80 reactors guarantee construction and operation simplicity (Chang et al. 2017; Merchuket 81 al. 2007), provide a high mass and heat transfer efficiency, and present low operating 82 costs (Vo et al. 2018; Zhang et al. 2018). 83 4 This research comparatively investigated the TMA removal performance of a bacterial 84 bubble column bioreactor (BC) and an algal-bacterial bubble column photobioreactor 85 (PBC), with a special focus on the quality of the treated gas (TMA and CO2 86 concentrations) and liquid effluent (concentration of N-containing species). The 87 structure of the microbial community in the PBC was also analyzed by pyrosequencing. 88 89 2 Materials and Methods 90 2.1 Inoculum 91 Activated sludge from Valladolid wastewater treatment plant (Valladolid, Spain) was 92 used to inoculate the BC, while a mixed inoculum containing activated sludge and 93 microalgae (1:1 v/v) was employed for the inoculation of the PBC. The microalgae were 94 obtained from a biogas upgrading high rate algal pond located at the Department of 95 Chemical Engineering and Environmental Technology at the University of Valladolid 96 (Valladolid, Spain) and operating with a total suspended solids (TSS) concentration of 97 1.62 g L-1 and a volatile suspended solids concentration (VSS) of 1.48 g L-1 (Franco98 Morgado et al. 2017). 99 100 2.2 Chemicals and mineral salt medium 101 The mineral salt medium (MSM) was composed of (g L-1): Na2HPO4·12H2O, 6.15; 102 KH2PO4, 1.52; MgSO4·7H2O, 0.2; CaCl2, 0.038; and 10 mL L-1 of a SL4 solution 103 containing (g L-1): EDTA, 0.5; FeSO4·7H2O, 0.2; ZnSO4·7H2O, 0.01; MnCl2·7H2O, 104 0.003. All the chemicals used for the preparation of the MSM were purchased in 105 Panreac (Barcelona, Spain). Trimethylamine (45 % purity) was obtained from Sigma 106 Aldrich (San Luis, EEUU). 107 5 2.3 Experimental setup and operating procedure 108 The experimental setup (Fig. 1) consisted of two cylindrical PVC columns (height = 109 0.58 m; inner diameter = 0.094 m) with a working volume of 4 L. The synthetic 110 contaminated emission was prepared by injecting the TMA liquid solution with a 111 syringe pump (Fusion 100, Chemyx Inc. USA) into an air stream of 2 L min-1, resulting 112 in an average inlet concentration of 513 ± 69 mg m-3 in each bioreactor. The gas stream 113 entered a mixing chamber in order to ensure complete TMA evaporation and 114 homogenization before being fed to the reactors through a porous diffuser (pore 115 diameter of 10 μm) located at the bottom. 116 For the inoculation of the BC, 2 L of aerobic activated sludge were centrifuged for 10 117 min at 10000 rpm and the pellet was resuspended in 1 L of MSM. The inoculum was 118 added to the BC and fresh MSM was supplemented upon filling the 4 L of working 119 volume, resulting in TSS and VSS concentrations of 2.79 and 2.13 g L-1, respectively. 120 The BC was operated for 78 days at an empty bed residence time (EBRT) of 2 min and 121 a daily replacement of 250 mL of the culture broth with fresh MSM (equivalent to a 122 dilution rate of 0.0625 d-1). During the first 50 days of operation, all the biomass was 123 recovered from the retrieved cultivation broth by centrifugation and returned to the 124 bioreactor (equivalent to an infinite solids retention time) in order to promote biomass 125 accumulation until reaching ~ 3 g VSS L-1. From day 50 onwards, 75 mL of the 250 mL 126 of the cultivation broth daily retrieved were discarded (cell retention time = 53.3 days) 127 in order to maintain a constant VSS concentration in the bioreactor. 128 A 1 h abiotic test was performed with no biomass prior PCB start-up. For this purpose, 129 the PCB illuminated with the LED lights was filled with mineral medium and TMA was 130 continuously supplied at an inlet concentration of ~700 mg m-3. The results 131 demonstrated that no TMA was removed by either adsorption or photodegradation 132 6 (Supplementary materials, Fig. S1). The inoculation of the PBC was performed by 133 centrifugation of 1 L of aerobic activated sludge and 1 L of microalgae culture (10 min, 134 10000 rpm). The pellets were resuspended in 1 L of MSM, added to the PBC and filled 135 up to 4 L with fresh MSM at initial TSS and VSS concentrations of 2.18 and 1.76 g L-1, 136 respectively. The PBC was operated for 103 days. CO2 was added to the inlet TMA137 laden emission at a concentration of 6 % v/v in order to supply inorganic carbon for 138 photosynthetic microalgae growth. To this end, 1.88 L min-1 of TMA-laden air were 139 mixed with 0.12 L min-1 of pure CO2 (Abelló Linde, Spain). A set of LED was installed 140 around the PBC, providing a photosynthetic active radiation (PAR) of ~ 250 µmol m-2 s141 1 at the outer reactor surface. The PBC was operated during the first 54 days at an 142 EBRT of 2 min with a daily MSM exchange rate of 250 mL (equivalent to a dilution 143 rate of 0.0625 d-1) (Stage I). Between days 55 and 79, the EBRT was reduced to 1.5 min 144 and the MSM exchange rate increased up to 375 mL d-1 (Stage II). Finally, from day 80 145 onwards, the EBRT was further decreased to 1 min and 500 mL of MSM were daily 146 exchanged (Stage III). During the first 12 days of operation, the biomass was recovered 147 from the withdrawn cultivation broth and returned to the PBC after centrifugation in 148 order to increase VSS concentration in the reactor. From this day on, the amount of 149 biomass returned to the system was adjusted in order to maintain a constant biomass 150 concentration of 3.5 g VSS L-1. 151 7 152 Fig. 1 Schematic representation of the experimental setup. GS: Gas sampling port 153 2.4 Analytical procedure 154 TMA gas concentration was analyzed in a Bruker 3900 gas chromatograph (Palo Alto, 155 USA) equipped with a flame ionization detector and a Supelco HP-5-MS (30 m × 0.25 156 µm × 0.25 µm) column. The oven, detector and injector temperatures were maintained 157 constant at 250, 300 and 200 °C, respectively, for 2.5 min. N2 was used as the carrier 158 gas at a flow of 1 mL min−1. CO2 and O2 gas concentrations were determined in a 159 Bruker 430 gas chromatograph (Palo Alto, USA) coupled with a thermal conductivity 160 detector and equipped with a CP-Molsieve 5A (15 m × 0.53 µm × 15 µm) and a P161 PoraBOND Q (25 m × 0.53 µm × 10 µm) columns. Oven, detector and injector 162 temperatures were maintained constant at 45, 200 and 150 °C for 5 min, respectively. 163 Helium was used as the carrier gas at a flow of 13.7 mL min−1. The pressure in the inlet 164 gas stream was daily measured using a differential pressure sensor IFM (Essen, 165 Germany) in order to control the actual flow of the inlet gas into the reactor. 166 Air Mixing chamber Flow controller Gas outlet (Photo) Bubble Column Reactor GS CO2 Syringe Pump MSM exchange GS 8 The pH was daily analyzed in the cultivation broth using a glass membrane electrode 167 PH BASIC 20 (Crison, Barcelona, Spain). Dissolved oxygen (DO) and temperature 168 were also analyzed in the cultivation broth of the PBC using a CellOx 325 oxygen meter 169 with a temperature sensor (WTW, New York, EEUU). Samples of the liquid phase of 170 both bioreactors were drawn twice a week for the determination of TSS, total organic 171 carbon (TOC), inorganic carbon (IC), total nitrogen (TN), ammonia (NH4+), NO2and 172 NO3concentrations. TSS and VSS concentrations were determined according to 173 standard methods (American Water Works Association, 2012). TOC, IC and TN 174 concentrations were measured using a TOC-VCSH analyzer coupled with a TNM-1 175 chemiluminescence module (Shimadzu, Japan). NH4+ concentration was analyzed with 176 an Orion Dual Star ammonium specific electrode (Thermo Scientific, The Netherlands). 177 Finally, 1 mL samples of cultivation broth were filtered through 0.22 μm filters and 178 analyzed by HPLC-IC for nitrite and nitrate determination using a Waters 515 HPLC 179 pump coupled with a conductivity detector (Waters 432) and equipped with an IC-PAK 180 Anion HC column (4.6 × 150 mm) and an IC-Pak Anion Guard-Pak (Waters). Samples 181 were eluted isocratically at 2 mL min−1 (at room temperature) with a solution of distilled 182 water/acetonitrile/n-butanol/buffer at 84/12/2/2% v/v (Muñoz et al. 2013). The 183 determination of the elemental composition of the algal-bacterial biomass in terms of 184 carbon (C), hydrogen (H) and nitrogen (N) content was conducted at the end of the 185 experimental period using a LECO CHNS-932 analyzer. 186 187 2.5 DNA extraction, illumina library preparation and pyrosequencing 188 Two samples were drawn for biological analysis from the cultivation broth of the PBC: 189 I-PBC (corresponding to the algal-bacterial inoculum) and F-PBC (at the end of the 190 experimental period). Total genomic DNA was extracted from 500 µL of sample using 191 9 the Fast DNA Spin kit for soil (Biomedical, USA) according to the manufacturer’s 192 instructions. DNA concentration was estimated by the Qubit fluorometer from 193 Invitrogen, and the final concentration of the DNA sample was normalized to 5 ng µL-1. 194 The extracted DNA was stored at -20ºC prior to pyrosequencing. Amplicon sequencing 195 was carried out targeting the 16S V3 and V4 regions (464bp, Escherichia coli based 196 coordinates) with the bacterial primers S-D-Bact-0341-b-S-17 and S-D-Bact-0785-aA197 21, forward and reverse, respectively, which were chosen according to (Klindworth et 198 al. 2013). Illumina adapter overhang nucleotide sequences were added to the gene199 specific sequences, thus resulting in the following full-length primers for the analysis: 200 5´TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCA 201 G (16S amplicon PCR forward primer), and 5´ 202 GTCTCGTGGGCTCGGAGATGTGTATAAGAGAC203 AGGACTACHVGGGTATCTAATCC (16S amplicon PCR reverse primer). Indexed 204 paired-end libraries were generated using the Nextera XT DNA Sample Preparation Kit 205 (Illumina, San Diego, CA), with a reduced number of PCR cycles (25) using 55 ºC as 206 annealing temperature. Libraries were then normalized and pooled prior to sequencing. 207 Non-indexed PhiX library (Illumina, San Diego, CA) was used as performance control. 208 Samples containing indexed amplicons were loaded onto the MiSeq reagent cartridge 209 and onto the instrument along with the flow cell for automated cluster generation and 210 paired-end sequencing with dual s (2 × 300bp run, MiSeq Reagent Kit v3) (Illumina, 211 San Diego, CA). The pyrosequencing analysis was carried by the Foundation for the 212 Promotion of Health and Biomedical Research of Valencia Region (FISABIO, Spain). 213 214 215 216 16 0.61 g L-1, in stages I, II and III, respectively. It is important to remark that a higher 347 biomass growth was recorded in the PBC compared to the BC due to the contribution of 348 the algal biomass. 349 350 3.3 Comparative analysis between BC and PBC 351 Overall, the PBC showed a better TMA removal performance than the conventional BC 352 at an EBRT of 2 min, with EC ×1.3 times higher compared to those recorded in the BC 353 at this EBRT. This improved behavior was attributed to the higher pH prevailing in the 354 PBC, which remained close to optimum values (6-8) for the enzymatic activity of 355 TMA-degrading bacteria (Chang et al. 2004). In this context, the pH in BC remained 356 below this optimal interval likely due to nitrification, while an average value of 6.7 ± 357 0.3 was recorded in the PCB due to N assimilation by microalgae and the inherent 358 increase in pH caused by photosynthesis. These favorable environmental conditions 359 allowed reducing the EBRT in the PCB to 1 min without statistically significant 360 differences in the RE and higher ECs (up to ×2.3 higher), despite the increase in TMA 361 load. This fact was attributed to a mass transfer limitation rather than a biological 362 limitation in the photobioreactor. In this sense, an increase in the TMA load resulted in 363 a higher concentration gradient and therefore an enhanced TMA mass transfer from the 364 gas to the liquid phase, where TMA-degrading microorganisms were capable of 365 sustaining the removal performance of the system. 366 Likewise, an improvement in the quality of the liquid effluent in terms of N content 367 under comparable TMA removal efficiencies was observed in the PBC. A total nitrogen 368 mass balance showed that 30% less nitrogen was discharged in the exchanged PBC 369 cultivation broth compared to that of the BC, even at EBRTs of 1.5 and 1 min (when 370 TMA load was 1.5 and 2 times higher, respectively) (Fig. 4). TN concentration 371 17 decreased significantly from 1307 ± 149 mg N L-1 in the BC to 879 ± 58, 893 ± 39 and 372 886 ± 78 mg N L-1 in the PBC in stages I, II and III, respectively. This decrease was 373 associated to nitrogen assimilation during algal biomass growth. Indeed, a biomass 374 production of ~0.23 and 0.70 g biomass d-1 was recorded in the BC and the PBC, 375 respectively. In this context, nitrogen is the most abundant macronutrient in algal 376 biomass with a content ranging between 5 and 10 % of its dry weight, as confirmed by 377 the analysis of CHN content of algal-bacterial biomass (42.3 ± 4.2 % C, 6.0 ± 0.6 % H 378 and 6.2 ± 1.3 % N). Nitrogen can be assimilated in the forms of NO3ˉ, NO2ˉ, NO or 379 NH4+, although assimilation of NH4+ over nitrite and nitrate is preferred by microalgae 380 as a result of its most reduced redox state (Markou et al. 2014). Interestingly, NO2ˉ 381 concentration increased in the PBC during stage I while an increase in NH4+ 382 concentration was recorded during stage II. This was attributed to the different activity 383 of the bacteria involved in the nitrification process depending on the operating 384 conditions of the PBC, resulting in the inhibition of the different stages of nitrification, 385 where ammonia oxidizing bacteria (AOB) oxidize NH4+ to NO2ˉ, which is subsequently 386 oxidized to NO3ˉ by nitrite-oxidizing bacteria (NOB). In this regard, the accumulation 387 of nitrite in stage I was attributed to the partial nitrification of NH4+ as a result of the 388 high temperature in the reactor (~31 ºC), which could hinder the activity of NOB such 389 as Nitrobacter (optimum temperature range ~ 24–25 °C) (Huang et al. 2010; Awolusi et 390 al. 2016). Thus, an incomplete nitrification would trigger the accumulation of NO2ˉ in 391 the medium. On the other hand, NH4+ concentration increased in stage II. The higher 392 TMA load applied at this lower EBRT might have inhibited nitrifying bacteria activity 393 due to their greater sensitivity to changes in NH4+ loading rates (Hu et al. 2009; Awolusi 394 et al. 2016), thus preventing NH4+ nitrification. In stage III, NO2ˉ concentration 395 18 significantly increased up to values close to those of stage I, probably due to the 396 acclimation of nitrifying bacteria to the temperature and NH4+ loading rates. 397 Overall, the optimal operating conditions were recorded in the PBC during Stage III 398 since similar values of TMA RE (> 90 %) and total nitrogen concentration in the 399 exchanged cultivation broth were recorded compared to Stages I and II, while the TMA 400 elimination capacity increased by ×2.3. 401 402 403 404 Fig. 4 Concentration profiles of nitrogen-containing species in the BC (white columns) and 405 PBC at the three EBRTs tested: (I) 2 min, (II) 1.5 min and (III) 1 min. Vertical lines represent 406 standard deviation. Columns within each group with different letters were significantly different 407 at p < 0.05 408 409 3.4 Effect of trimethylamine on the microbial communities 410 A total of 133521 and 105547 initial bacterial 16S rRNA sequence reads generated by 411 the MiSeq Illumina platform for I-PBC and F-PBC samples, respectively, passed the 412 quality and taxonomic cutoff. Effective bacterial sequences from the samples were 413 19 affiliated to a total of 11 phyla. Of them, the most dominant phyla in I-PBC were 414 Actinobacteria (32 %), Proteobacteria (17 %), Candidatus Saccharibacteria (15 %), 415 Chloroflexi (11 %) and Firmicutes (9 %). Other phyla with abundances > 1 % were 416 Bacteroidetes, Planctomycetes and Verrucomicrobia. These phyla are commonly found 417 in activated sludge (Zhang et al. 2012; Lebrero et al. 2013). However, a significant 418 specialization of the microbial community was observed as a result of TMA 419 biodegradation, which resulted in the dominance of only two phyla: Actinobacteria (55 420 %) and Proteobacteria (42 %) (Fig. 5). Indeed, the Shannon-Wiener diversity indices of 421 the microbial communities present at the I-PBC and F-PBC were 3.68 and 1.86, 422 respectively. Typical values range from 1.5 to 3.5, which correspond to low and high 423 species evenness and richness, respectively (MacDonald 2003). The significant decrease 424 in the diversity index revealed a gradual enrichment and specialization of the microbial 425 community as a result of TMA biodegradation. 426 427 Fig. 5 Community composition at a phyla level across samples. I-PBC: algal-bacterial 428 inoculum, F-PBC: end of the experimental period. The abundance is presented in terms of 429 percentage in total effective bacterial sequences in a sample, classified using RDP Classifier 430 431 20 At the genus level, the pyrosequencing showed a total of 501 genera in I-PBC and 139 432 in F-PBC sample (Fig. 6). Of them, only 82 in I-PBC and 20 in F-PBC were present 433 with abundances > 0.1 % (Table S1), and represented 95 and 98 % of the total number 434 of readings at the genus level, respectively, which confirmed the lower diversity in the 435 F-PBC. Based on the phylogenetic analysis, the Actinobacteria phylum was the most 436 abundant in both samples, the genus Mumia, belonging to the family Nocardioidaceae, 437 representing 47 % of the total genera. The Mumia genus was recently discovered in soil 438 samples (Lee et al. 2014), thus growth conditions and functions are yet unknown. The 439 results obtained in the present study suggest the ability of Mumia to grow on TMA, 440 although further research is necessary to confirm this hypothesis. 441 442 21 443 Fig. 6 Krona graphs showing the population structure of samples I-PBC (A: algal-bacterial 444 inoculum) and F-PBC (B, end of the experimental period) 445 446 The genus Aquamicrobium belonging to the class Alphaproteobacteria showed an 447 abundance of 38 %. Different members of the genus Aquamicrobium have been isolated 448 from pollutant-loaded environments such as wastewater treatment plants, activated 449 sewage sludge and biofilters (Jin et al. 2013). Moreover, Aquamicrobium sp. was 450 recently identified as an AOB resistant to high concentrations of N-NH4+ (Yang et al. 451 2015). Similarly, Huang et al. demonstrated the capacity of Aquamicrobium sp. to 452 oxidize ammonia, and classified this genus as a coldand salt-tolerant AOB (Huang et 453 al. 2017). Therefore, the presence of Aquamicrobium in the PBC cultivation broth, an 454 AOB extremely tolerant to high temperatures and NH4+ concentrations, supports the 455 hypothesis of a partial nitrification which resulted in the accumulation of nitrite over 456 nitrate. 457 22 458 4. Conclusions 459 This study confirmed the feasibility of biologically abating TMA from waste gas 460 streams in bacterial and algal-bacterial bubble column reactors. The bacterial reactor 461 achieved REs of 78 % and ECs of 12 g TMA m-3 h-1 at inlet TMA concentrations of ~ 462 500 mg m-3 and EBRT of 2 min. Conversely, the algal-bacterial photobioreactor 463 provided enhancements in TMA removal by almost 20 % and reached ECs of 16 g 464 TMA m-3 h-1 under similar conditions. The maintenance of high TMA-REs at EBRTs of 465 1.5 and 1 min (98 % and 94 %, respectively) confirmed the outstanding performance of 466 the algal-bacterial photobioreactor. The higher pH recorded in the PBC due to 467 photosynthetic activity together with the lower nitrification rates could have mediated 468 this enhanced performance. Moreover, algal activity in the PBC resulted in a net CO2 469 consumption in the gas stream and a 30 % decrease in the TN concentration of the 470 liquid effluent as a result of a superior nitrogen assimilation. These promising results 471 highlight the potential of implementation of this innovative process for TMA abatement 472 from air emissions. Likewise, this study supports the relevance of future research in 473 order to adapt the process to the specific needs of the industrial emissions. 474 475 Acknowledgements 476 This work was supported by the regional government of Castilla y León and the EU477 FEDER programme (UIC 71 and CLU 2017-09). 478 479 480 481 23 References 482 Aguirre A, Bernal P, Maureira D, et al (2018) Biofiltration of trimethylamine in biotrickling 483 filter inoculated with Aminobacter aminovorans. Electron J Biotechnol 33:63–67. doi: 484 10.1016/j.ejbt.2018.04.004 485 American Water Works Association, 2012. 486 Standard Methods for the Examination of Water and Wastewater. American Water Works 487 Association/American Public 488 Awolusi OO, Nasr M, Kumari S, Bux F (2016) Artificial Intelligence for the Evaluation of 489 Operational Parameters Influencing Nitrification and Nitrifiers in an Activated Sludge 490 Process. Microb Ecol 72:49–63. doi: 10.1007/s00248-016-0739-3 491 Borde X, Guieysse B, Delgado O, et al (2003) Synergistic relationships in algal-bacterial 492 microcosms for the treatment of aromatic pollutants. Bioresour Technol 86:293–300. doi: 493 10.1016/S0960-8524(02)00074-3 494 Chang CT, Chen BY, Shiu IS, Jeng FT (2004) Biofiltration of trimethylamine-containing waste 495 gas by entrapped mixed microbial cells. Chemosphere 55:751–756. doi: 496 10.1016/j.chemosphere.2003.11.037 497 Chang J-S, Show P-L, Ling T-C, et al (2017) Photobioreactors. Curr Dev Biotechnol Bioeng. 498 doi: 10.1016/B978-0-444-63663-8.00011-2 499 Cole JR, Wang Q, Cardenas E, et al (2009) The Ribosomal Database Project: Improved 500 alignments and new tools for rRNA analysis. Nucleic Acids Res. doi: 10.1093/nar/gkn879 501 Ding Y, Wu W, Han Z, Chen Y (2008) Correlation of reactor performance and bacterial 502 community composition during the removal of trimethylamine in three-stage biofilters. 503 Biochem Eng J 38:248–258. doi: 10.1016/j.bej.2007.07.011 504 Edgar RC (2010) Search and clustering orders of magnitude faster than BLAST. 505 Bioinformatics. doi: 10.1093/bioinformatics/btq461 506 Estrada JM, Kraakman NJRB, Muoz R, Lebrero R (2011) A Comparative Analysis of Odour 507 Treatment Technologies in Wastewater Treatment Plants. Environ Sci Technol 45:1100– 508 1106. doi: 10.1021/es103478j 509 Franco-Morgado M, Alcántara C, Noyola A, et al (2017) A study of photosynthetic biogas 510 upgrading based on a high rate algal pond under alkaline conditions: Influence of the 511 illumination regime. Sci Total Environ 592:419–425. doi: 10.1016/j.scitotenv.2017.03.077 512 Ho KL, Chung YC, Lin YH, Tseng CP (2008) Biofiltration of trimethylamine, dimethylamine, 513 24 and methylamine by immobilized Paracoccus sp. CP2 and Arthrobacter sp. CP1. 514 Chemosphere 72:250–256. doi: 10.1016/j.chemosphere.2008.01.044 515 Hu J, Li DP, Liu Q, et al (2009) Effect of organic carbon on nitrification efficiency and 516 community composition of nitrifying biofilms. J Environ Sci 21:387–394. doi: Doi 517 10.1016/S1001-0742(08)62281-0 518 Huang X, Bai J, Li K ran, et al (2017) Characteristics of two novel coldand salt-tolerant 519 ammonia-oxidizing bacteria from Liaohe Estuarine Wetland. Mar Pollut Bull 114:192– 520 200. doi: 10.1016/j.marpolbul.2016.08.077 521 Huang Z, Gedalanga PB, Olson BH (2010) Distribution of <I>Nitrobacter</I> and 522 <I>Nitrospira</I> Communities in an Aerobic Activated Sludge Bioreactor and their 523 Contributions to Nitrite Oxidation. Proc Water Environ Fed 2010:2390–2403. doi: 524 10.2175/193864710798159101 525 Jin HM, Kim JM, Jeon CO (2013) Aquamicrobium aestuarii sp. nov., a marine bacterium 526 isolated from a tidal flat. Int J Syst Evol Microbiol 63:4012–4017. doi: 527 10.1099/ijs.0.048561-0 528 Kang D, Zhao Q, Wu Y, et al (2017) Removal of nutrients and pharmaceuticals and personal 529 care products from wastewater using periphyton photobioreactors. Bioresour Technol. doi: 530 10.1016/j.biortech.2017.06.153 531 Kim SG, Bae HS, Oh HM, Lee ST (2003) Isolation and characterization of novel halotolerant 532 and/or halophilic denitrifying bacteria with versatile metabolic pathways for the 533 degradation of trimethylamine. FEMS Microbiol Lett 225:263–269. doi: 10.1016/S0378534 1097(03)00530-5 535 Klindworth A, Pruesse E, Schweer T, et al (2013) Evaluation of general 16S ribosomal RNA 536 gene PCR primers for classical and next-generation sequencing-based diversity studies. 537 Nucleic Acids Res. doi: 10.1093/nar/gks808 538 Lebrero R, Volckaert D, Pérez R, et al (2013) A membrane bioreactor for the simultaneous 539 treatment of acetone, toluene, limonene and hexane at trace level concentrations. Water 540 Res. doi: 10.1016/j.watres.2013.01.041 541 Lee LH, Zainal N, Azman AS, et al (2014) Mumia flava gen. nov., sp. nov., an actinobacterium 542 of the family Nocardioidaceae. Int J Syst Evol Microbiol 64:1461–1467. doi: 543 10.1099/ijs.0.058701-0 544 Liffourrena AS, Lucchesi GI (2014) Degradation of trimethylamine by immobilized cells of 545 Pseudomonas putida A (ATCC 12633). Int Biodeterior Biodegrad 90:88–92. doi: 546 10.1016/j.ibiod.2014.02.008 547 25 MacDonald G (2003) Biogeography: Introduction to Space, Time, and Life. Prof Geogr. doi: 548 10.1111/0033-0124.5502018 549 Magoč T, Salzberg SL (2011) FLASH: Fast length adjustment of short reads to improve 550 genome assemblies. Bioinformatics. doi: 10.1093/bioinformatics/btr507 551 Markou G, Vandamme D, Muylaert K (2014) Microalgal and cyanobacterial cultivation: The 552 supply of nutrients. Water Res 65:186–202. doi: 10.1016/j.watres.2014.07.025 553 Merchuk JC, Garcia-Camacho F, Molina-Grima E (2007) Photobioreactor design and fluid 554 dynamics. Chem Biochem Eng Q 21:345–355. 555 Muñoz R, Guieysse B (2006) Algal–bacterial processes for the treatment of hazardous 556 contaminants: A review. Water Res 40:2799–2815. doi: 10.1016/j.watres.2006.06.011 557 Muñoz R, Souza TSO, Glittmann L, et al (2013) Biological anoxic treatment of O2-free VOC 558 emissions from the petrochemical industry: A proof of concept study. J Hazard Mater 559 260:442–450. doi: 10.1016/j.jhazmat.2013.05.051 560 Oksanen J, Blanchet FG, Kindt R, et al (2015) vegan: Community Ecology Package. R package 561 version 2.0-10. 2013. R Packag ver 24–3. doi: 10.1007/s10265-013-0586-y 562 Ondov BD, Bergman NH, Phillippy AM (2011) Interactive metagenomic visualization in a Web 563 browser. BMC Bioinformatics. doi: 10.1186/1471-2105-12-385 564 Oyarzun P, Alarcón L, Calabriano G, et al (2019) Trickling filter technology for biotreatment of 565 nitrogenous compounds emitted in exhaust gases from fishmeal plants. J Environ Manage 566 232:165–170. doi: 10.1016/j.jenvman.2018.11.008 567 Perillo PM, Rodríguez DF (2016) Low temperature trimethylamine flexible gas sensor based on 568 TiO2 membrane nanotubes. J Alloys Compd 657:765–769. doi: 569 10.1016/j.jallcom.2015.10.167 570 Schmieder R, Edwards R (2011) Quality control and preprocessing of metagenomic datasets. 571 Bioinformatics. doi: 10.1093/bioinformatics/btr026 572 Vo HNP, Bui XT, Nguyen TT, et al (2018) Effects of nutrient ratios and carbon dioxide bio573 sequestration on biomass growth of Chlorella sp. in bubble column photobioreactor. J 574 Environ Manage 219:1–8. doi: 10.1016/j.jenvman.2018.04.109 575 Wan S, Li G, Zu L, An T (2011) Bioresource Technology Purification of waste gas containing 576 high concentration trimethylamine in biotrickling filter inoculated with B350 mixed 577 microorganisms. Bioresour Technol 102:6757–6760. doi: 10.1016/j.biortech.2011.03.059 578 Wang Q, Garrity GM, Tiedje JM, Cole JR (2007) Naïve Bayesian classifier for rapid 579 assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 580