Salinity is the major driver of the global eukaryotic community structure in fish-canning wastewater treatment plants

Salinity is the major driver of the global eukaryotic community structure in fish-canning wastewater treatment plants David Correa-Galeote, Alba Roibás, Anuska Mosquera-Corral, Belén Juárez-Jiménez, Jesús González-López, Belén Rodelas Accepted manuscript How to cite: Journal of Environmental Management, 290 (2021), 112623. https://doi.org/10.1016/j.jenvman.2021.112623 Copyright information: © 2021 Elsevier Ltd. This manuscript version is made available under the CC-BY-NC-ND 4.0 license (http://creativecommons.org/licenses/by-nc-nd/4.0/) 1 Title 1 Salinity is the major driver of the global eukaryotic community structure in fish-canning 2 wastewater treatment plants 3 David Correa-Galeote1,2, Alba Roibas3, Anuska Mosquera-Corral3, Belen Juárez-Jiménez1,2, Jesús 4 González-López1,2, Belen Rodelas1,2 5 1 Universidad de Granada. Facultad de Farmacia. Dpto. de Microbiología 6 2 Universidad de Granada. Instituto del Agua. Sección Microbiología y Tecnologías Ambientales 7 3 Universidade de Santiago de Compostela. Escuela de Ingeniería. Dpto. de Ingeniería Química 8 Email: [email protected] 9 10 11 Key Words: Eukarya community, Eukaryotic diversity, NaCl concentration, Fish-canning WWTP, 12 Illumina sequencing 13 14 Highlights 15 NaCl concentration was a main driver in the modulation of the eukaryotic diversity 16 A greater eukaryotic diversity seems required for saline wastewater treatment 17 A high relative abundance of dark matter fungi was found, which was related to NaCl 18 Ocystis (Chlorophyta) is proposed as an eukaryotic bioindicator in saline WWTPs 19 20 21 Abstract 22 Fish-canning wastewater is characterized frequently by a high content of salt (NaCl), making 23 its treatment particularly difficult; however, the knowledge of the effect of NaCl on eukaryotic 24 communities is very limited. In the present study, the global diversity of eukaryotes in activated 25 sludges (AS) from 4 different wastewater treatment plants (WWTPs) treating fish-canning 26 effluents varying in salinity (0.47, 1.36, 1.72 and 12.76 g NaCl/L) was determined by sequencing 27 partial 18S rRNA genes using Illumina MiSeq. A greater diversity than previously reported was 28 2 observed in the AS community, which comprised 37 and 330 phylum-like and genera-like groups, 29 respectively. In this sense, the more abundant genus-like groups (average relative abundance 30 (RA) >5%) were Adineta (6.80%), Lecane (16.80%), Dictyostelium (7.36%), Unclassified_Fungi7 31 (6.94%), Procryptobia (5.13) and Oocystis (5.07%). The eukaryotic communities shared a 32 common core of 25 phylum-like clades (95% of total sequences); therefore, a narrow selection 33 of the eukaryotic populations was found, despite the differences in the abiotic characteristics of 34 fish-canning effluents and reactor operational conditions inflicted. The differences in NaCl 35 concentration were the main factor that influenced the structure of the eukaryotic community, 36 modulating the RAs of the different phylum-like clades of the common core. Higher levels of salt 37 increased the RAs of Ascomycota, Chlorophyta, Choanoflagellata, Cryptophyta, Mollusca, 38 Nematoda, Other Protists and Unclassified Fungi. Among the different eukaryotic genera here 39 found, the RA of Oocystis (Chlorophyta) was intimately correlated to increasing NaCl 40 concentrations and it is proposed as a bioindicator of the global eukaryotic community of fish41 canning WWTPs. 42 43 Abbreviations 44 AS: Activated sludge 45 CAS: conventional activate sludge 46 NMS: nonmetric multidimensional scaling 47 OTU: Operational taxonomic unit 48 PCA: Principal component analysis 49 RA: Relative abundance 50 TOC: Total organic carbon 51 TSS: Total suspended solids 52 VFA: Volatile fatty acids 53 VSS: Volatile suspended solids 54 3 WW: Wastewater 55 WWTP: Wastewater treatment plants 56 57 1. Introduction 58 World population growth has corresponded with an increase in the global consumption per 59 capita of fish and seafood, which has doubled since 1973 with >3% average annual increase 60 (FAO, 2018). Nowadays, fish and seafood represent about 17% of the world's animal protein 61 intake (Laso et al., 2018). The proportion of fish processed into prepared or preserved forms 62 equals 10% of total human fish consumption, amounting >15 million tonnes (live weight) per 63 year globally. Among the different preservation methods, canning transformation is of great 64 importance, as it improves shelf life enabling storage for several years. In the Iberian Peninsula, 65 the canning sector is the main segment of fishing manufacturing industry with >85 factories, 66 located mainly in the coastal areas (Val del Rio et al., 2018). 67 The fish canning industry consumes large amounts of water for cleaning and washing raw 68 materials, boiling and cooking processes, resulting in equally large quantities of wastewater 69 (WW) (Lim et al., 2003). Fish canning effluents require treatment before their disposal to prevent 70 environmental damage, but they are characterized generally by a high content of salt and 71 organic matter (including oil and grease), which makes their treatment particularly difficult (Val 72 del Rio et al., 2018). These processes are performed in wastewater treatment plants (WWTPs), 73 which usually include primary units aimed to remove particulate matter and oils by 74 physicochemical treatments, and in some cases secondary units that normally consist of 75 biological reactors for the removal of dissolved matter (Pham et al., 2020). 76 In secondary treatments, Conventional Activated Sludge (CAS) systems are robust and 77 widely employed to treat both urban and industrial wastewater because of their versatility. 78 Besides, the combination of aerobic and anoxic periods or basins promotes the removal of 79 nutrients and organic matter together in the same system. Nevertheless, this WWTP approach 80 4 includes only treatment, and no valorisation is performed. For systems including anaerobic 81 digestion, organic matter is converted into biogas. However, nutrients are not removed in the 82 valorisation unit (the anaerobic digester), so an extra unit is necessary (normally, a CAS or an 83 anammox unit) where nutrients and/or the remaining organic matter are removed. Moreover, 84 high salinity is known to inhibit methanogenesis, lowering biogas yield (Palmeiro-Sánchez et al., 85 2013). Therefore, the use of this type of treatment can be challenging in high-salinity systems as 86 fish-canning WWTP. Novel wastewater treatment technologies include the use of biofilms or 87 membranes. Although the results obtained are promising (Carrera et al., 2019), the full-scale 88 application of these technologies is still scarce, and most of the industrial fish-canning WWTPs 89 operate by using a CAS or an anaerobic digester. 90 The biological processes taking place in the secondary treatments are based on the activity 91 of a consortium of microbial populations. Among the different microorganisms, eukaryotes are 92 pivotal players in WWTP bioreactors (dos Santos et al., 2014; Hirakata et al., 2019). In this sense, 93 some genera have been described as beneficial in pollutant removal, such as members of the 94 Fungi, Chlorophyta, or Rhizaria (an unranked group), which contribute to reduce nitrogen and 95 phosphorus concentrations in WW (Zahedi et al., 2019). Also, phyla Ciliophora, Rotifera and 96 Nematoda, or Stramenopiles (unranked group), are bacterivores that regulate the proliferation 97 of bacterial populations (Šimek et al., 2019). However, some eukaryotic genera have also been 98 reported as harmful for technical or health management reasons (Zahedi et al., 2019), pointing 99 out the importance of microbial communities’ identification. 100 High-throughput sequencing analysis of 18S rRNA gene has revolutionised the 101 understanding of eukaryotes diversity and abundance in WWTPs (Cooper et al., 2016). Recently, 102 some efforts have been made to characterize specific branches of the tree of life of the Eukarya 103 domain by using these tools (Hirakata et al., 2019; Liu et al., 2019). Despite these attempts, the 104 knowledge of the biodiversity of different eukaryotic communities in WWTP is still very limited 105 and unidentified eukaryotes with important roles may be overlooked. 106 5 Different physicochemical parameters affect the eukaryotic communities in WWTPs, among 107 which salinity might be a key factor for their composition and robustness (Horikoshi et al., 2010). 108 However, little is known about the effect of NaCl in these communities in fish-canning WWTPs. 109 As effluents coming from fish-canning industries are characterized by variable salinity 110 concentrations (2-35 g NaCl/L, Cristóvão et al., 2016), it is essential to comprehensively 111 characterize their global eukaryotic communities and to establish how salinity modulates the 112 dynamics of the populations that comprise them. 113 In order to go deeper into the existing relationships among the different eukaryotic clades, 114 and aiming to establish how operational parameters and WW influent characteristics drive the 115 biodiversity of those communities, several goals are expected to be fulfilled in the present work: 116 (1) to determine the global eukaryotic biodiversity in fish-caning WWTP bioreactors; (2) to 117 investigate the relationships among the different eukaryotic communities; (3) to examine the 118 effect that different physicochemical parameters, mainly salinity, produce in the composition of 119 the Eukarya community; and (4) to propose bioindicator genera of eukaryotes whose abundance 120 correlates to the shifts of different operational parameters in fish-canning WWTPs. 121 The answers to these objectives will expand the knowledge of the diversity of the eukaryotic 122 communities in the biological treatment of high-salinity effluents, helping to infer their potential 123 roles. These results will be also relevant for the biological treatment of other saline effluents, 124 such as those derived from chemical, pharmaceutical, agricultural, or aquacultural industries 125 (Zhao et al. 2020). 126 127 2. Experimental procedures 128 2.1 Description of the WWTPs source of biomass samples 129 Duplicate sets of samples (n=2), referred to as F1 to F4, were taken in four different WWTPs 130 of fish-canning industries in Galicia (Northwest of Spain) under steady-state conditions. F1 and 131 F2 were CAS systems operating at low (0.47g NaCl/L) and moderate (1.72 g NaCl/L) salinities, 132 6 respectively, while F3 (moderate salinity, 1.36 g NaCl/L) and F4 (high salinity level, 12.76 g 133 NaCl/L) valorise organic matter by converting it into biogas in anaerobic units before using CAS 134 for nutrient removal (Fig. 1). 135 2.2 Chemical determinations 136 Collected liquid samples were characterized as elsewhere described (Val del Rio et al., 2018; 137 Pedrouso et al., 2021). The pH value was measured with a glass electrode (Crisson GLP22). Total 138 Suspended Solids (TSS) and Volatile Suspended Solids (VSS) were analysed, in bulk samples, 139 according to the Standard Methods (APHA, 2005). Samples were filtered using a cellulose-ester 140 filter (Advantec, Japan) when soluble parameters were determined. The concentrations of 141 sodium (Na+) and chloride (Cl-) were analysed by ion chromatography (Metrohm 861, 142 Switzerland). Volatile fatty acids (VFA) concentrations were measured by gas chromatography 143 (6850 Series II, Agilent Technologies). Total organic carbon (TOC) concentration was determined 144 by a Shimazu analyser (TOC-L, automatic sample injector Shimadzu ASI-L). Ammonia was 145 measured spectrophotometrically at 640 nm (Bower and Holm-Hansen, 1980). 146 2.3 DNA isolation 147 Different volumes of activated sludge (AS) samples (9 - 13.5 mL) were centrifuged at 14,000 148 rpm at 22 °C for 1 min, the supernatants were discarded and biomass kept at -20 °C until use. 149 Two independent biological replicates were used for each DNA extraction, using the FastDNA 150 Spin kit for Soil in the FastPrep-24 instrument (MP Biomedicals, Irvine, CA, USA) according to the 151 manufacturer’s instructions, and stored at -20 °C until further use. 152 2.4 Next-generation sequencing library preparation 153 DNA extracts were provided to RTL Genomics (Lubbock, Texas, USA) for partial 18S rRNA 154 Illumina MiSeq sequencing. PCR amplification products of the 18S rRNA V9 region gene were 155 generated using the Earth Microbiome Project primers Euk1391f (5′‐GTACACACCGCCCGTC‐3′) 156 (Lane, 1991) and EukBr (5′‐TGATCCTTCTGCAGGTTCACCTAC‐3′) (Medlin et al. 1988). As the 157 forward primer is a highly-conserved three domain‐level primer, there is a risk of recovery of 158 7 non‐eukaryotic tags even when used in combination with an eukaryote‐specific PCR primer 159 (Cooper et al., 2016). However, the highly conserved nature of the eukaryote‐specific PCR 160 primer Euk1391f was reported to recover higher diversity, compared to other alternative 161 primers designed specifically for eukaryotes amplification only (Amaral‐Zettler et al., 2009). 162 2.5 Bioinformatic analysis of 18S rRNA gene amplicon libraries 163 Raw Illumina sequencing data of industrial WWTPs were analysed using the Mothur v.1.42.3 164 pipeline (Schloss et al., 2009). Full details of the followed procedures are provided in 165 Supplementary material, Section 2. 166 2.6 Sequencing coverage estimation and α-diversity indices 167 Good's coverage index and Simpson, Shannon and Chao-1 indices were calculated according 168 to Esty (1986) and Hill et al. (2003), respectively. 169 2.7 Statistical analyses 170 Analyses of variance of the different parameters among the samples were made using the 171 non-parametric Kruskal-Wallis and Conover-Iman combined tests, with a 95% significance level 172 (p<0.05) using the package XLSTAT v2019.3.2 (Addinsoft, New York, USA). Heatmaps were 173 constructed using the average clustering method, and generated with R studio v.3.4.1 package 174 (Rstudio, Boston, MA, USA). Principal component analysis (PCA) and nonmetric 175 multidimensional scaling (NMS) ordination analysis were constructed by using PC‐ORD software 176 (Wild Blueberry Media, Oregon, USA). Full details are described in Supplementary materials. 177 Additionally, correlations analyses of each one of the abiotic variables and population RAs at the 178 genus-like level were made using Spearman's rank correlation coefficients, using the XLSTAT 179 software. 180 2.8 Nucleotide sequence accession numbers 181 The nucleotide sequences reported in this work have been deposited in GeneBank under 182 the accession number SUB7560150. 183 184 8 3. Results 185 3.1 Wastewater and sludge characterization 186 WW and AS samples were characterized in order to determine the different conditions to 187 which microorganisms were exposed. The results displayed in Table 1 show the high variability 188 of fish-canning effluents. As previously stated, salinity was low in F1, moderate in F2 and F3, and 189 high for F4. F4 is a high-load system, with a TOC concentration in the influent of approximately 190 2 g/L, while the remaining facilities treat low or moderate organic loads (0.26-0.49 g/L). 191 The PCA plot of the physico-chemical properties explains 87% of the total variability among 192 the 4 WWTPs (Fig. S1). The samples were distributed around all the 2D-space created, indicating 193 a high variability of the different abiotic parameters. NaCl concentration in WW was the main 194 factor driving the ordination of the samples, as it was the variable more strongly related to the 195 principal component 1. 196 3.2 Next-generation sequencing library summary 197 Microbial community structures of AS retrieved from the four WWTPs were determined by 198 Illumina sequencing. Overall, a total of 164,128 high-quality sequences were obtained, 199 corresponding to 5,148 unique sequences. Posteriorly, sequences were de novo clustered into 200 3,070 operational taxonomic units (OTUs, 3% divergence). On average, 64.38% of the 201 representative sequences of each OTU were classified within the Eukarya domain, while 31.50% 202 were related to Bacteria, and 3.81% were not classified into any domain (Fig. 2). Unclassified 203 sequences and sequences classified as Bacteria were filtered and excluded, and 106,158 204 sequence reads were retained and used for subsequent analyses. 205 3.3 Overall eukaryotic community structure 206 The total number of eukaryotic OTUs was 1,413 (Table S1). In order to facilitate the analysis 207 of the composition and structure of the eukaryotic community, the different clades were 208 grouped according to their corresponding kingdom-like groups, as detailed in Supplementary 209 material (Section 2.1). The relative abundance (RA) of the kingdom-like groups was 27.74% for 210 15 were placed apart independently. Analogous results of similitude among samples were observed 366 in the heatmap of the 37 phylum-like taxa (Fig. 5). Blastocladiomycota, Chytridiomycota, 367 Cnidaria, Euglenozoa, Gastrotricha, Microsporidia, Mucoromycota, Platyhelminthes, Rhizaria, 368 Rhodophyta, Rotifera and Unclassified_eukaryotes, were grouped around the F1 and F2 sample 369 scores. Similarly, Annelida, Amoebozoa, Aphelida, Apicomplexa, Bacillariophyta, Basidiomycota, 370 Cercozoa, Ciliophora, Heterolobosea, Jakobida, Porifera, Streptophyta and Zoopagomycota 371 clustered around the ordination of F3 samples. Finally, F4 amplicon libraries were mainly related 372 to Alveolata, Arthropoda, Ascomycota, Choanoflagellata, Chordata, Chlorophyta, Cryptophyta, 373 Mollusca, Nematoda, Other_Prostists, Stramenopiles and Unclassified_Fungi. 374 The vectors representing the physicochemical variables of the WW and sludges from the 375 different WWTPs were overlapped on the NMS ordination (Fig. 7). Pairwise Pearson product376 moment (r) correlations between the shifts of the different abiotic variables and the RAs of the 377 37 phylum-like clades shown in Fig. 7 are detailed in Table S3. Absolute values of r >0.7 were 378 considered as strong correlations. The pH in WW and AS, TOC in WW, and NaCl concentration 379 in WW and AS, were positively related to Alveolata, Arthropoda, Ascomycota, Chlorophyta, 380 Choanoflagellata, Chordata, Chytridiomycota, Cryptophyta, Mollusca, Nematoda, 381 Other_Protists, Stramenopiles, Unclassified_eukaryotes and Unclassified_Fungi. On the other 382 hand, NH4+ and VFA concentrations in WW, and VSS concentration in AS, were highly related to 383 Amoebozoa, Annelida, Aphelida, Apicomplexa, Bacillariophyta, Basidiomycota, Cercozoa, 384 Ciliophora, Heterolobosea, Jakobida, Porifera, Streptophyta and Zoopagomycota. In contrast, 385 the content of NH4+ in the WW was negatively related to Alveolata, Arthropoda, Chordata, 386 Chytridiomycota, Cnidaria, Stramenopiles and Unclassified_eukaryotes. Finally, the VFA 387 concentration in WW and VSS concentration in AS were negatively correlated to Alveolata, 388 Arthropoda, Chordata, Chytridiomycota, Cnidaria, Mucoromycota and Unclassified_eukaryotes. 389 3.7 Proposal of eukaryotic taxa as bioindicators in fish-canning WWTPs 390 16 Finally, pairwise Spearman's rank correlations between the abiotic variables and the RAs of 391 taxa at the genus-like level were calculated, and a high number of significant correlations were 392 found (Table S4). The ten strongest positive or negative correlations are displayed in Tables 3 393 (abiotic properties of WW) and 4 (abiotic properties of AS). pH in WW and NaCl content in both 394 WW and AS were highly correlated among them (Fig. S1); therefore, these variables correlated 395 similarly with the RAs of the same genus-like clades. Parallel trends were also observed for TOC 396 in WW and VSS concentration in AS, while the rest of the physico-chemical properties were 397 correlated to biotic data independently. Despite different genus-like clades were intimately 398 related to NaCl concentration, it is worth mentioning that the RA of Oocystis had the maximum 399 level of correlation with salinity in both WW and AS (Tables 3 and 4). This genus-like taxon may 400 play a pivotal role in the eukaryotic community of the analysed WWTPs, since it was within the 401 common core and reached high RAs in all the samples (0.9-20.0%); therefore, it is proposed as 402 a bioindicator of the response of eukaryotic microbial communities to the shifts of NaCl 403 concentration in fish-canning treatment systems. 404 405 4. Discussion 406 4.1 Community structure at the kingdom-like level 407 In this work, 259-329 eurakyotic OTUs were detected in AS samples (Table 2), with no 408 statistical differences of their richness among the four WWTPs. In contrast, a much lower 409 number of eukaryotic OTUs has previously been recorded in conventional WWTPs (García‐Ruíz 410 et al., 2020) and WWTPs treating sewage from chemical industries (Zhao et al., 2019). Analysis 411 of the biodiversity at the phylum-like level allowed us to differentiate 37 groups (Table S2). In 412 contrast, Chouari et al. (2017) described 16 different phylum-like groups by cloning and 413 sequencing DNA isolated from different parts of an urban WWTP, while García-Ruiz et al. (2020) 414 reported only 19 phylum-like taxa when they analysed the biodiversity in the anaerobic digester 415 of a pilot-scale WWTP bioreactor. Therefore, compared to the previous observations in other 416 17 WWTPs, the results presented here revealed a greater diversity of the eukaryotic communities 417 in AS of fish-canning WWTP and, similarly, suggest that the treatment of the high-strength 418 and/or saline influents entering the fish-canning treatment facilities required a higher number 419 of eukaryotic groups than in other WWTPs. 420 4.2 Eukaryotic biodiversity synchrony 421 The synchrony of the eukaryotic biodiversity among the 4 WWTPs, determined as the 422 number of shared phylum-like clades present in each one of the different AS samples (Fig. 4 and 423 Table S3), was noteworthy. Therefore, the variations either in physic-chemical parameters or 424 the operational design of the WWTPs did not result in a drastic modification of the eukaryotic 425 diversity at the phylum-like level, but rather modulated the community structure (Fig. 4 and 426 Table S2). Besides, the OTUs richnesses and the values of the Shannon and Chao-1 indices were 427 not significantly different among samples (Table S2). In contrast, a lower level of synchrony has 428 been previously described in the eukaryotic composition of WWTPs by other authors. Zahedi et 429 al. (2019) reported that only 22% of sequences at the phylum-like level were shared in 4 rural 430 WWTPs. Similarly, Matsunaga et al. (2014) indicated that the abundance of shared eukaryotic 431 OTUs in sludges from different WWTPs was around 10%; however, in this work the common 432 core was composed by 34% of the identified OTUs. This suggests that the characteristics of fish433 canning influents produce a stronger selection of the eukaryotic communities than that 434 observed in other types of WWTPs. 435 The likenesses among the AS samples were also analysed globally taking into account the 436 RAs of taxa at the phylum and genus-like levels through construction of heatmaps. Considered 437 generally, there was a higher level of similarity between the F1 and F2 samples than for the 438 other two WWTPs (Fig. 4 and Figs. S2c, S3c, S4c, S5c and S6c), except for the fungal communities, 439 wherein the higher level of similarity was observed between F2 and F3, both bearing moderate 440 salinity. Taking into account that the F1 and F2 WWTPs have the same configuration, the use of 441 CAS systems resulted in a more specific selection of the eukaryotic community than that 442 18 observed for F3 and F4 facilities. Besides, the NMS ordination based on the RA of the 37 phylum443 like clades confirmed this result (Fig. 7). Therefore, the type of configuration influenced the 444 development of the eukaryotic community structures of the 4 WWTPs, in agreement with 445 previous reports (Matsunaga et al., 2014). 446 The high level of synchrony of the eukaryotic communities among the 4 WWTPs was also 447 supported by the different correlation analyses described in Fig. 6 and Figs. S7-S9. Also, it is 448 important to note that the different interactions were not circumscribed to a specific kingdom, 449 as in each one of the 3 clusters, members of the 5 kingdom-like clades analysed were found. It 450 has been previously reported that a high level of relationships among eukaryotes is important 451 for the adequate maintenance of the ecological equilibrium of the food-web and the nutrient 452 cycling in a given ecosystem (Peng et al., 2018). Consequently, the high level of co-occurrence 453 patterns observed in this work suggests that a highly imbricated structure of the eukaryotic 454 community is required for an effective WW treatment in this type of industrial WWTPs. These 455 results agree with those previously described by dos Santos et al. (2014), who analysed 456 protozoan and metazoan diversity in mixed WW (70% industrial and 30% domestic influents). 457 Also, it is remarkable the absence of human pathogens such as Entamoeba (Amoebozoa), 458 Giardia (Fornicata) or Enterobius (Nematoda) among the dominant genera in the AS. Similar 459 results were observed in intermediate treatment stages and effluent samples of different 460 domestic WWTPs by Zahedi et al. (2919), while other authors reported a high presence of these 461 pathogens in urban WWTPs effluents (Ma et al. 2016). 462 4.3 Roles of the core phylum-like taxa in WWTPs 463 Salinity increases are usually linked to a worsening in the system performance (Pham and 464 Nguyen, 2020; Cristóvão et al., 2016). However, the driving mechanisms of this common 465 behaviour are generally unknown, and the effect of salinity over eukaryotic populations, which 466 play a key role in wastewater treatment, has been rarely studied. Due to the scarce number of 467 previous scientific works evaluating the relationships between the performance of WWTPs and 468 19 the eukaryotic diversity of their AS, potential roles of the core phylum-like taxa are proposed in 469 this work mostly according to their life-styles described in other ecosystems. 470 The shared phylum-like clades within the paraphyletic group of Protista, in which 471 relationships remain poorly understood in terms of diversity and systematics (Santoferrara et 472 al., 2017), were Alveolata, Amoebozoa, Ciliophora, Euglenozoa and Other_Protists. In general, 473 protists are ubiquitous microorganisms (Galal et al., 2018) of central importance in the 474 functioning of microbial food-webs, where they play diverse trophic roles (McManus and 475 Santoferrara, 2013) and subsequently, they are of vital importance in WWTPs. The members of 476 the supergroup Alveolata, which includes a large range of trophic modes and habitats, act as 477 important marine primary producers (Guillou et al., 2008). Amoebozoa comprises organisms 478 with different feeding behaviours (bacterivory, detritivory, predatory and mixotrophic) 479 (Macumber et al., 2020), and several previous works described its occurrence in WWTPs 480 (Chouari et al., 2017; Cohen et al., 2019). Ciliophora, also known as ciliates, are major 481 environmental micropredators with grazing activity that could directly stimulate bacterial 482 colonies (Johansson et al. 2004), and have been regarded as a main eukaryotic group in different 483 WWTPs (Peng et al., 2018; Šimek et al., 2019). Euglenozoa or flagellates have been described in 484 AS from WWTPs (dos Santos et al., 2014; Yoshino et al., 2017), where their occurrence could be 485 linked to their heterotrophic and phagotrophic life-style. 486 Fungal taxa shared in the samples analyzed this study were the phylum-like groups 487 Ascomycota, Blastocladiomycota, Microsporidia, Unclassified_Fungi, and other minority groups. 488 Nowadays, it is well established that fungal communities are prime movers in WWTPs as they 489 are potentially important contributors to various functions (Niu et al., 2017; Wang et al., 2017). 490 In addition to their beneficial roles for sewage treatment, fungi can be also involved in bulking 491 and foaming phenomena, resulting in poor settling and dewatering of AS in some cases (Hossain, 492 2004). Ascomycota display a high diversity of life-styles (Aranda, 2016) and they can contribute 493 to organic compounds' biodegradation, the aggregation of sludge flocs and detoxification (Niu 494 20 et al., 2017); accordingly, their biodiversity and abundance are of great importance in sewage 495 treatment. Ascomycota was recognized as the most abundant fungal phyla in urban WWTPs in 496 several previous studies (Maza-Márquez et al., 2016; Niu et al., 2017; Assress et al., 2019); 497 however, in this work, the prevalence of this phylum was minoritarian. Blastocladiomycota 498 includes saprotrophs or parasites of plants, animals or other fungi, and they usually account for 499 a minor fraction of the fungal community in WWTP bioreactors (Maza-Márquez et al., 2016; 500 Zhang et al., 2019), as also observed here. 501 Remarkably, a large quantity of fungal sequences remained unclassified at the phylum-like 502 level and were part of the dark matter fungi (Fig. 4 and Table S2), whose biodiversity and ecology 503 are largely unknown (Ryberg and Nilsson, 2018). A high correlation between NaCl concentration 504 and the RAs of dark matter fungi was revealed, which has not been previously reported in 505 industrial WWTPs. In this sense, the F4 facility, which bore the highest salinity, was particularly 506 enriched in unclassified fungi (37% of the overall eukaryotic community). Extreme environments 507 harbour a wider content of uncharacterized microorganisms whose occurrence indicates many 508 undiscovered microbial niches and interactions (Nobu et al., 2015). Therefore, there is still a gap 509 in the knowledge of the fungal diversity and population interactions in fish-canning WWTP, and 510 the relevance of the links among fungal population shifts and salinity needs to be further 511 explored. 512 Inside the Metazoa kingdom, Arthropoda, Cnidaria, Nematoda, Platyhelminthes, Rotifera 513 and other minority phyla (Annelida, Chordata, Mollusca and Porifera) were shared among the 4 514 WWTPs. The different phyla within the Metazoa could be responsible for the improvement of 515 WWTP performance by controlling the density of bacteria by predation, reducing sludge 516 production, and enhancing sludge flocculation (Wang et al., 2017). Arthropoda are members of 517 the eukaryotic community of full-scale WWTPs (Ju et al., 2014; Bedoya et al. 2019), where they 518 play a major role in the food-webs in two ways, by decomposing organic material for themselves, 519 and by providing substrates for other eukaryotes via cracking through internal digestion 520 21 (Dehghani et al., 2016). The majority of Cnidaria members have predatory life-styles, but certain 521 species may also scavenge dead animals (Frazão et al., 2012). Their presence and roles in WWTP 522 bioreactors are not fully stablished but could be linked to the performance of a control over the 523 microbial populations. Nematoda and Platyhelminthes include microbotrophics (ingesting 524 bacteria, algae, single-celled fungi and protozoa) or predaceous lifestyles on other eukaryotes 525 (Noreña, et al., 2015). Several works have earlier described the presence of both phyla in AS 526 from different urban and industrial WWTPs (Leal et al., 2013; El Fels et al., 2019). The highly 527 abundant phylum Rotifera encompasses predatory or microphagous organisms (Wallace et al., 528 2015). They are a promising tool for controlling the over-proliferation of filamentous bacteria in 529 AS diminishing sludge bulking (Pajdak-Stós et al. 2017), and they are often found in industrial 530 WWTPs (dos Santos et al., 2014). The minority phyla Annelida, Chordata, Mollusca and Porifera, 531 are mainly related to organic carbon content reduction in sewage treatment (Zhao et al., 2010), 532 and the occurrence and roles of most of them are not yet well-established in WWTPs. 533 The shared phyla from the Viridiplantae and related kingdom-like rank were Chlorophyta, 534 Rhodophyta, Stramenopiles and Streptophyta. Chlorophyta have a high potential for 535 improvement of WWTP performance, being able to reduce organic matter, remove N and P 536 nutrients and heavy metals (Ariesyady et al., 2016; Schulze et al., 2017; Peralta et al., 2019), and 537 generate O2 through photosynthesis (Chen et al., 2017). It is not surprising that the abundance 538 of this phylum was highly correlated to salt concentrations (Table S3 and Fig. 7) as they naturally 539 occur in seawater, indicating that their members are halotolerants and can be pivotal players in 540 WWTPs subjected to high salinity sewage, as it is the case of fish-canning WWTPs. Rhodophyta 541 usually are employed as microhabitats for diverse invertebrates and other algae (Aguirre et al., 542 2017) and their presence in WWTPs could theoretically increase the diversity of eukaryotes. The 543 existence of Rhodophyta in WWTPs is scarcely reported, as only the work of Ghosh and Love 544 (2011) previously found them abundant in WWTPs. This shows the highly specific biodiversity of 545 the community found in the analysed WWTPs, probably due to the particular conditions of fish546 22 canning influents. Stramenopiles can act as main grazers of bacteria and play key roles within 547 the microbial food-web. Although the presence of this group in WWTP bioreactors has been well 548 demonstrated previously (Chouari et al., 2017; Cohen et al, 2019), in this work, only unclassified 549 sequences were found and, subsequently, a description of their hypothetical roles cannot be 550 inferred. 551 Finally, 6 minority phylum-like clades (Cercozoa, Choanoflagellata, Heterolobosea, 552 Jakobida, Rhizaria and Unclassified_eukaryotes, Fig. 4 and Table S2) were grouped together in 553 the hodgepodge Other Eukaryotes kingdom-like group. The shared phylum-like clades included 554 Cercozoa, Heterolobosea, Rhizaria and Unclassified_eukaryotes. The presence of both Rhizaria 555 and Cercozoa were previously described at the different steps of the treatment process in 556 WWTPs (Peng et al., 2018; Hirakata et al., 2019), where they are able to reduce N and P 557 concentrations in WW (Zahedi et al., 2019). Besides, the members of Rhizaria are mostly 558 predatory heterotrophs (Cavalier-Smith et al., 2018). Most heteroloboseans are bacterivores; 559 also, they are able to feed on different eukaryotes such as diatoms, or even on themselves in 560 a cannibalistic manner. Although they have been related to hypersaline environments (Park 561 et al., 2007), no robust correlations between their abundance and NaCl content were found in 562 this work (Table S3). Some of the Unclassified eukaryotes identified in the AS samples have 563 been recently proposed as predators (Ancoracysta, Janouškovec et al., 2017) or grazers 564 (Telonema, Yabuki et al., 2013, and Katablephari, Kwon et al., 2017). The presence of these taxa 565 in WWTPs was only previously confirmed for the genus Anurofeca (He et al., 2017). 566 4.4 Effect of NaCl on eukaryotic biodiversity 567 The PCA (Fig S1) pointed out a stronger influence of the NaCl concentration in both WW and 568 AS over other abiotic variables and, similarly, the role of salt stress on the shaping of the 569 eukaryotic community structure was demonstrated in the NMS analysis (Fig. 7). Therefore, NaCl 570 concentration was a major driver of the shifts in the structure of the eukaryotic community. In 571 this sense, the changes in salinity resulted in a modulation of the RAs of the different phylum572 23 like clades rather than a change in the phyla that formed the eukaryotic community (Table 3 and 573 4). Recently, Rodríguez-Sánchez et al. (2019) reported that community structures of Eukarya 574 were affected by the salinity level more than by other factors and were more responsive to its 575 changes than the prokaryotic communities in two hybrid moving-bed pilot scale bioreactors and 576 a membrane bioreactor treating urban WW. Similarly, other studies showed that salinity was a 577 driving factor for the shaping of the eukaryotic community in different aquatic ecosystems 578 (Rojas-Jiménez et al., 2019). On the other hand, although several authors highlighted the 579 negative effect of NaCl on the abundance of different phylum-like clades (Horikoshi et al., 2010; 580 Cortés-Lorenzo et al., 2016; Rodríguez-Sánchez et al., 2019), strikingly, no strong negative 581 correlations were observed in this work (Fig. 7, Table S4). Taking into account the origin of the 582 raw material treated in this type of WWTP, it is not surprising that the eukaryotic populations 583 were well adapted to cope with salinity. In this sense, different mechanisms to protect 584 eukaryotic cells of the negative effect of NaCl have been developed by eukaryotes (Fuentes et 585 al, 2016; Harding and Simpson, 2018; Vashishtha and Meghwanshi, 2018). Besides, Pham and 586 Nguyen (2020) previously reported that the efficiency of CAS systems was significantly affected 587 under NaCl contents >3% due to cell lysis. The results presented here indicate that a broader 588 eukaryotic diversity could reduce the negative effect of NaCl and promote the colonization of 589 the different ecological niches. 590 4.5 Bioindicator genera in fish-canning WWTPs 591 Several authors have proposed different eukaryotic genera as bioindicators of an optimal 592 performance of WWTP bioreactors; i.e., members of the Breviatea (genus Subulatomonas), and 593 Chlorophyta (genus Desmodesmus) (Chouari et al., 2017; He et al., 2017; Hirakata et al., 2019). 594 Similarly, the occurrence of the genus Arcella (Amoebozoa), Geotrichum (Ascomycota) and 595 Tokophrya (Ciliphora) have been associated with deterioration of WWTP effluent quality (dos 596 Santos et al., 2014; Zhang et al., 2016). However, none of the abovementioned genera 597 accounted for a significant fraction of the eukaryotic communities in the fish-canning treatment 598 24 samples analysed here. Therefore, it is necessary to propose eukaryotic genera that can be used 599 as bioindicators in industrial WWTPs that operate under high salinity conditions. 600 As earlier stated in section 2.7, Oocystis (phylum Chlorophyta) is a core genus-like clade in 601 the 8 amplicon libraries here analysed, and its RA was strongly correlated to increasing values 602 of NaCl concentrations of both WW (Table S4). Oocystis are microscopic coccoid green algae 603 often described in many aquatic ecosystems (Štenclová et al., 2017), and Huang et al. (2012) 604 described members of this genus as primary populations that efficiently reduce the 605 concentration of dissolved N compounds in different saltwater systems. Another advantage of 606 the presence of Oocystis in WWTPs is their ability to tolerate high levels of heavy metals (Soldo 607 and Behra 2000). Therefore, Oocystis may contribute to the maintenance of the ecological 608 equilibrium of the food-web and the nutrient cycling in WW subjected to high concentrations of 609 NaCl or heavy metals. Therefore, the high prevalence and ubiquity of Oocystis across the 4 fish610 canning WWTPs and its enrichment under increasing NaCl concentrations suggests that this 611 genus is a good candidate to be proposed as bioindicator in WWTPs of fish-canning facilities. 612 This work was pioneer in the analysis of global eukaryotic diversity in fish-canning WWTP. 613 Taking into account all data, the results here described significantly contribute to improve the 614 understanding of the ecology and putative functions of the different members of the eukaryotic 615 communities in industrial WWTPs operated under varying levels of salinity, since studies related 616 to this topic are currently scarce. In general, the putative contributions of eukaryotes into the 617 food-web could be acting as decomposers, primary producers or as predators/grazers of 618 bacteria, small metazoans and other eukaryotes. Other works are necessary in order to confirm 619 the specific contributions of different genera to the optimal operation of these industrial 620 WWTPs. Gathering diversity information from this and future work will help to elucidate the 621 ecology of the different members of the domain Eukarya and enable the proposal of more 622 bioindicator taxa useful to evaluate the operational state of industrial WWTPs. Despite the 623 advances in the knowledge of the ecology and metabolic roles of eukaryotes in sewage 624 31 Huang, X., Li, X., Wang, Y., Zhou, M., 2012. Effects of environmental factors on the uptake rates 781 of dissolved nitrogen by a salt-water green alga (Oocystis borgei snow). B. Environ. 782 Contam. Tox. 89, 905–909. https://doi.org/10.1007/s00128-012-0767-8 783 Janouškovec, J., Tikhonenkov, D.V., Burki, F., Howe, A.T., Rohwer, F.L., Mylnikov, A.P., Keeling, 784 P.J., 2017. A new lineage of eukaryotes illuminates early mitochondrial genome reduction. 785 Cur. Biol. 27, 3717–3724. https://doi.org/10.1016/j.cub.2017.10.051 786 Johansson, M., Gorokhova, E., Larsson, U.L.F., 2004. Annual variability in ciliate community 787 structure, potential prey and predators in the open northern Baltic Sea proper. J. Plankton 788 Res. 26, 67–80. https://doi.org/10.1093/plankt/fbg115 789 Ju, F., Guo, F., Ye, L., Xia, Y., Zhang, T., 2014. Metagenomic analysis on seasonal microbial 790 variations of activated sludge from a full‐scale wastewater treatment plant over 4 years. 791 Environ. Microbiol. Rep. 6, 80–89. https://doi.org/10.1111/1758-2229.12110 792 Kwon, J.E., Jeong, H.J., Kim, S.J., Jang, S.H., Lee, K.H., Seong, K.A., 2017. Newly discovered role 793 of the heterotrophic nanoflagellate Katablepharis japonica, a predator of toxic or harmful 794 dinoflagellates and raphidophytes. Harmful Algae. 68, 224–239. 795 https://doi.org/10.1016/j.hal.2017.08.009 796 Lane, D.J., 1991. 16S/23S rRNA sequencing, in: Stackebrandt, E., Goodfellow, M. (Eds.), Nucleic 797 acid techniques in bacterial systematics. Wiley, New York, pp. 115–175. 798 Laso, J., García-Herrero, I., Margallo, M., Vázquez-Rowe, I., Fullana, P., Bala, A., et al., 2018. 799 Finding an economic and environmental balance in value chains based on circular 800 economy thinking: An eco-efficiency methodology applied to the fish canning industry. 801 Resour. Conserv. Recycl. 133, 428–437. https://doi.org/10.1016/j.resconrec.2018.02.004 802 Leal, A.L., Dalzochio, M.S., Flores, T.S., de Alves, A.S., Macedo, J.C., Valiati, V.H., 2013. 803 Implementation of the sludge biotic index in a petrochemical WWTP in Brazil: improving 804 operational control with traditional methods. J. Ind. Microbiol. Biotechnol. 40, 1415– 805 1422. https://doi.org/10.1007/s10295-013-1354-7 806 32 Lim, L.C., Dhert, P., Sorgeloos, P., 2003. Recent developments and improvements in ornamental 807 fish packaging systems for air transport. Aquac. Res. 34, 923–935. 808 https://doi.org/10.1046/j.1365-2109.2003.00946.x 809 Liu, T., He, J., Cui, C., Tang, J., 2019. Exploiting community structure, interactions and functional 810 characteristics of fungi involved in the biodrying of storage sludge and beer lees. J. 811 Environ. Manage. 232, 321–329. https://doi.org/10.1016/j.jenvman.2018.11.089 812 Ma, J., Feng, Y., Hu, Y., Villegas, E.N., Xiao, L., 2016. Human infective potential of Cryptosporidium 813 spp., Giardia duodenalis and Enterocytozoon bieneusi in urban wastewater treatment 814 plant effluents. J. Water Health. 14, 411–423. https://doi.org/10.2166/wh.2016.192 815 Macumber, A.L., Blandenier, Q., Todorov, M., Duckert, C., Lara, E., Lahr, D.J., et al., 2020. 816 Phylogenetic divergence within the Arcellinida (Amoebozoa) is congruent with test size 817 and metabolism type. Eur. J. Protistol. 72, 125645. 818 https://doi.org/10.1016/j.ejop.2019.125645 819 Matsunaga, K., Kubota, K., Harada, H., 2014. Molecular diversity of eukaryotes in municipal 820 wastewater treatment processes as revealed by 18S rRNA gene analysis. Microbes 821 Environ. 29, 401–407. https://doi.org/10.1264/jsme2.ME14112 822 Maza-Márquez, P., Vilchez-Vargas, R., Kerckhof, F.M., Aranda, E., González-López, J., Rodelas, B., 823 2016. Community structure, population dynamics and diversity of fungi in a full-scale 824 membrane bioreactor (MBR) for urban wastewater treatment. Water Res. 105, 507–519. 825 https://doi.org/10.1016/j.watres.2016.09.021 826 McManus, G.B., Santoferrara, L.F., 2013. Tintinnids in microzooplankton communities, in: Dolan, 827 J.R., Montagnes, D.J., Agatha, S., Coats, D.W., Stoecker, D.K. (Eds.), Biology and Ecology of 828 Tintinnid Ciliates: ciliates: models for marine plankton. Wiley-Blackwell, Oxford, pp. 198– 829 213. 830 33 Medlin, L., Elwood, H.J., Stickel, S., Sogin, M.L., 1988. The characterization of enzymatically 831 amplified eukaryotic 16S-like rRNA-coding regions. Gene. 71, 491–499. 832 https://doi.org/10.1016/0378-1119(88)90066-2 833 Niu, L., Li, Y., Xu, L., Wang, P., Zhang, W., Wang, C., Cai, W., et al., 2017. Ignored fungal 834 community in activated sludge wastewater treatment plants: diversity and altitudinal 835 characteristics. Environ. Sci. Pollut. R. 24, 4185–4193. https://doi.org/10.1007/s11356836 016-8137-4 837 Nobu, M.K., Narihiro, T., Rinke, C., Kamagata, Y., Tringe, S.G., Woyke, T., Liu, W.T., 2015. 838 Microbial dark matter ecogenomics reveals complex synergistic networks in a 839 methanogenic bioreactor. ISME J. 9, 1710–1722. https://doi.org/10.1038/ismej.2014.256 840 Noreña, C., Damborenea, C., Brusa, F., 2015. Phylum Platyhelminthes, in: Thorp, J.H., Rogers, 841 D.C. (Eds.), Ecology and General Biology: Freshwater Invertebrates. Academic Press, 842 London, pp. 181–203. 843 Pajdak-Stós, A., Kocerba-Soroka, W., Fyda, J., Sobczyk, M., Fiałkowska, E., 2017. Foam-forming 844 bacteria in activated sludge effectively reduced by rotifers in laboratory-and real-scale 845 wastewater treatment plant experiments. Environ. Sci. Pollut. Res. 24, 13004–13011. 846 https://doi.org/10.1007/s11356-017-8890-z 847 Palmeiro-Sánchez, T., Val del Río, A., Mosquera-Corral, A., Campos, J.L., Méndez, R., 2013. 848 Comparison of the anaerobic digestion of activated and aerobic granular sludges under 849 brackish conditions. Chem. Eng. J. 231, 449–454. 850 https://doi.org/10.1016/j.cej.2013.07.052 851 Park, J.S., Simpson, A.G., Lee, W.J., Cho, B.C., 2007. Ultrastructure and phylogenetic placement 852 within Heterolobosea of the previously unclassified, extremely halophilic heterotrophic 853 flagellate Pleurostomum flabellatum (Ruinen 1938). Protist. 158, 397–413. 854 https://doi.org/10.1016/j.protis.2007.03.004 855 34 Pedrouso, A., Correa-Galeote, D., Maza-Márquez, P., Juárez-Jimenez, B., González-López, J., 856 Rodelas, B., et al., 2021. Understanding the microbial trends in a nitritation reactor fed 857 with primary settled municipal wastewater. Sep. Purif. Technol. 256, 117828. 858 https://doi.org/10.1016/j.seppur.2020.117828 859 Peng, R.H., Xiong, A.S., Xue, Y., Fu, X.Y., Gao, F., Zhao, W., et al., 2008. Microbial biodegradation 860 of polyaromatic hydrocarbons. FEMS Microbiol. Rev. 32, 927–955. 861 https://doi.org/10.1111/j.1574-6976.2008.00127.x 862 Peng, Y., Li, J., Lu, J., Xiao, L., Yang, L., 2018. Characteristics of microbial community involved in 863 early biofilms formation under the influence of wastewater treatment plant effluent. J. 864 Environ. Sci. 66, 113–124. https://doi.org/10.1016/j.jes.2017.05.015 865 Peralta, E., Jerez, C.G., Figueroa, F.L., 2019. Centrate grown Chlorella fusca (Chlorophyta): 866 Potential for biomass production and centrate bioremediation. Algal Res. 39, 101458. 867 https://doi.org/10.1016/j.algal.2019.101458 868 Pham, T., Nguyen, T., 2020. A study to use activated sludge anaerobic combining aerobic for 869 treatment of high salt seafood processing wastewater. Curr. Chem. Let. 9, 79–88. 870 https://doi.org/10.5267/j.ccl.2019.8.002 871 Rodríguez-Sánchez, A., Leyva-Diaz, J.C., Muñoz-Palazon, B., Poyatos, J.M., González-López, J., 872 2019. Influence of salinity cycles in bioreactor performance and microbial community 873 structure of membrane-based tidal-like variable salinity wastewater treatment systems. 874 Environ. Sci. Pollut. R. 26, 514–527. https://doi.org/10.1007/s11356-018-3608-4 875 Rojas-Jimenez, K., Rieck, A., Wurzbacher, C., Jürgens, K., Labrenz, M., Grossart, H.P., 2019. A 876 salinity threshold separating fungal communities in the Baltic Sea. Front. Microbiol. 10, 877 680. https://doi.org/10.3389/fmicb.2019.00680 878 Ryberg, M., Nilsson, R.H., 2018. New light on names and naming of dark taxa. MycoKeys. 2018, 879 31. https://doi.org/10.3897/mycokeys.30.24376 880 35 Santoferrara, L.F., Alder, V.V., McManus, G.B., 2017. Phylogeny, classification and diversity of 881 Choreotrichia and Oligotrichia (Ciliophora, Spirotrichea). Mol. Phylogenet. Evol. 112, 12– 882 22. https://doi.org/10.1016/j.ympev.2017.03.010 883 Schloss, P.D., Westcott, S.L., Ryabin, T., Hall, J.R., Hartmann, M., Hollister, E.B., et al., 2009. 884 Introducing mothur: open-source, platform-independent, community-supported 885 software for describing and comparing microbial communities. Appl. Environ. Microbiol. 886 75, 7537–7541. https://doi.org/10.1128/AEM.01541-09 887 Schulze, P.S., Carvalho, C.F., Pereira, H., Gangadhar, K.N., Schüler, L.M., Santos, T.F., et al., 2017. 888 Urban wastewater treatment by Tetraselmis sp. CTP4 (Chlorophyta). Bioresour. Technol. 889 223, 175–183. https://doi.org/10.1016/j.biortech.2016.10.027 890 Šimek, K., Grujčić, V., Nedoma, J., Jezberová, J., Šorf, M., Matoušů, A., et al., 2019. Microbial 891 food webs in hypertrophic fishponds: Omnivorous ciliate taxa are major protistan 892 bacterivores. Limnol. Oceanogr. 64, 2295–2309. https://doi.org/10.1002/lno.11260 893 Soldo, D., Behra, R., 2000. Long-term effects of copper on the structure of freshwater periphyton 894 communities and their tolerance to copper, zinc, nickel and silver. Aquat. Toxicol. 47, 81– 895 189. https://doi.org/10.1016/S0166-445X(99)00020-X 896 Štenclová, L., Fučíková, K., Kaštovský, J., Pažoutová, M., 2017. Molecular and morphological 897 delimitation and generic classification of the family Oocystaceae (Trebouxiophyceae, 898 Chlorophyta). J. Phycol. 53, 1263–1282. https://doi.org/10.1111/jpy.12581 899 Val del Rio, A., Pichel, A., Fernández-González, N., Pedrouso, A., Fra-Vázquez, A., Morales, N., et 900 al., 2018. Performance and microbial features of the partial nitritation-anammox process 901 treating fish canning wastewater with variable salt concentrations. J. Environ. Manage. 902 208, 112–121. https://doi.org/10.1016/j.jenvman.2017.12.007 903 Vashishtha, A., Meghwanshi, G.K., 2018. Fungi inhabiting in hypersaline conditions: an insight, 904 in: Gehlot, P., Singh, J. (Eds.), Fungi and their role in sustainable development: current 905 perspectives. Springer, Singapore, pp. 449–465. 906 36 Wallace, R.L., Snell, T.W., Smith, H.A., 2015. Phylum rotifer, in: Thorp, J.H., Rogers, D.C. (Eds.), 907 Ecology and General Biology: Freshwater Invertebrates. Academic Press, London, pp. 908 225–271. 909 Wang, Q., Wei, W., Gong, Y., Yu, Q., Li, Q., Sun, J., Yuan, Z., 2017. Technologies for reducing 910 sludge production in wastewater treatment plants: state of the art. Sci. Total Environ. 587, 911 510–521. https://doi.org/10.1016/j.scitotenv.2017.02.203 912 Yabuki, A., Eikrem, W., Takishita, K., Patterson, D.J., 2013. Fine structure of Telonema subtilis 913 Griessmann, 1913: a flagellate with a unique cytoskeletal structure among eukaryotes. 914 Protist. 164, 556–569. https://doi.org/10.1016/j.protis.2013.04.004 915 Yoshino, H., Suenaga, T., Fujii, T., Hori, T., Terada, A., Hosomi, M., 2017. Efficacy of a high916 pressure jet device for excess sludge reduction in a conventional activated sludge process: 917 Pilot-scale demonstration. Chem. Eng. J. 326, 78–86. 918 https://doi.org/10.1016/j.cej.2017.05.084 919 Zahedi, A., Greay, T.L., Paparini, A., Linge, K.L., Joll, C.A., Ryan, U.M., 2019. Identification of 920 eukaryotic microorganisms with 18S rRNA next-generation sequencing in wastewater 921 treatment plants, with a more targeted NGS approach required for Cryptosporidium 922 detection. Water. Res. 158, 301–312. https://doi.org/10.1016/j.watres.2019.04.041 923 Zhang, Q., Hu, J., Lee, D.J., 2016. Aerobic granular processes: current research trends. Bioresour. 924 Technol. 210, 74–80. https://doi.org/10.1016/j.biortech.2016.01.098 925 Zhang, M., Yao, J., Wang, X., Hong, Y., Chen, Y., 2019. The microbial community in filamentous 926 bulking sludge with the ultra-low sludge loading and long sludge retention time in 927 oxidation ditch. Sci. Rep. 9, 1–10. https://doi.org/10.1038/s41598-019-50086-3 928 Zhao, L., Wang, Y., Yang, J., Xing, M., Li, X., Yi, D., Deng, D., 2010. Earthworm–microorganism 929 interactions: a strategy to stabilize domestic wastewater sludge. Water Res. 44, 2572– 930 2582. https://doi.org/10.1016/j.watres.2010.01.011 931 37 Zhao, Y.J., Sato, Y., Inaba, T., Aoyagi, T., Hori, T., Habe, H., 2019. Activated sludge microbial 932 communities of a chemical plant wastewater treatment facility with high-strength 933 bromide ions and aromatic substances. J. Gen. Appl. Microbiol. 21, 106–110. 934 https://doi.org/10.2323/jgam.2018.05.002 935 Zhao, Y., Zhuang, X., Ahmad, S., Sung, S., Ni, S.Q., 2020. Biotreatment of high-salinity 936 wastewater: current methods and future directions. World J. Microbiol. Biotechnol. 36, 937 37. https://doi.org/10.1007/s11274-020-02815-4 938 939 940 Figure Captions 941 Figure 1. Flowchart of the processes carried out in the four sampled WWTP for the treatment of fish942 canning effluents. a) F1, low salinity and conventional activated sludge (CAS) system, b) F2, moderate 943 salinity and CAS system, c) F3, moderate salinity and anaerobic reactor for biogas production and CAS 944 system for nitrogen removal, d) F4, high salinity and anaerobic reactor for biogas production and CAS 945 system for nitrogen removal. Thin arrows represent the water treatment line, thick arrows the sludge line 946 and dashed arrows gas lines. 947 Figure 2. Average relative abundances of operational taxonomic units (OTUs) assigned to domain level 948 from activate sludge samples retrieved from WWTPs treating fish-canning wastewater by Illumina high949 throughput sequencing (n = 2). Sequences reads that were not classified into any domain were labelled 950 as “Unclassified”. According to the Kruskal-Wallis and Conover-Iman tests (p < 0.05), different letters 951 indicates significant differences among samples for a given domain. 952 Figure 3. Average relative abundances of operational taxonomic units (OTUs) assigned to kingdom-like 953 level of the Eukaryotic domain in activate sludge samples (n=2) by Illumina high-throughput sequencing 954 of WWTPs treating fish-canning wastewater. 955 Figure 4. Relative abundances (represented by the areas of dots) of different phylum-like clades in 956 activated sludge samples (n=2) of WWTPs treating fish-canning wastewater identified by high957 throughput Illumina sequencing. Eukaryotic groups representing at least 0.5% of the total sequences or 958 at least a 0.25% for a given sample were shown. Other_Protist includes the sequences related to 959 Diplonemea class, Petalomonadida order, Liburna and Carpediemonas genera and unclassified protists; 960 Other_Fungi includes Basidiomycota, Chytridiomycota and Mucoromycota phyla; Other_Metazoa 961 includes Annelida, Chordata, Mollusca and Porifera phyla. Other_Eukaryotics include the sequences 962 belonging to Cercozoa, Rhodophyta and Bacillariophyta phyla, Aphelida and Rhizaria groups, and 963 Unclassified Eukaryotes. 964 Figure 5. Heatmap of the average clustering of the relative abundance of the phylum-like clades 965 identified in the DNA isolated from activated sludge samples (n=2) from WWTP bioreactors treating fish966 canning industrial wastewater identified by high-throughput Illumina sequencing. 967 38 Figure 6. Hierarchical Spearman correlation matrix diagram of the 37 different groups found in the 968 activated sludges samples (n=2) in WWTPs treating fish-canning wastewater identified by Illumina high969 throughput sequencing. 970 Figure 7. Nonmetric Multidimensional Scaling (NMS) analysis of the relative abundance eukaryotic 971 phylum-like taxa (relativized to the maximal taxon abundance) in activated sludges sample (n=2) retrieved 972 from fish-canning industrial WWTPs by Illumina sequencing and its linking to the main physicochemical 973 properties (log (X + 1) transformed) of influents and activated sludges of the WWT plants (volatile 974 suspended solids of activated sludge, VSS_AS; pH of wastewater, pH_WW; pH of activated sludge, pH_AS; 975 NaCl concentration in wastewaters, NaCl_WW; NaCl concentration in activated sludges, NaCl_AS; volatile 976 fatty acids in wastewater, VFA_WW; ammonium concentration in wastewater, NH4+_WW). Samples from 977 F1: triangles; F2: diamonds; F3: circles; F4 samples: squares. 978 979