Layout 1 INTRODUCTION The anthropogenic impact is significantly altering aquatic ecosystems (Vörösmarty et al., 2010; Elliott and Elliott, 2013), by means of a plethora of pressures that in- clude chemical pollution and wastewater discharge, eu- trophication, hypoxia, utilization of living resources, habitat destruction and climate change effects (Halpern et al., 2008). Aquatic sediments, acting as repositories of materials and solutes deposited from or diffused through the water column, accumulate chemical and biological contaminants (Ridgway and Shimmield, 2002). This holds true also for microbiological pollutants, including au- tochthonous and pathogenic microbes of fecal origin, that reach the aquatic environment by a variety of routes and can spread diseases to human and aquatic populations, with important socio-economic, sanitary and environmen- tal consequences (Stewart et al., 2008). A number of studies have so far investigated the pres- ence and distribution of fecal bacteria in aquatic sedi- ments. The majority of these studies have addressed the traditional fecal indicators, such as total coliforms, Es- cherichia coli and intestinal enterococci, which are world- wide used for assessments of aquatic ecosystems quality (Field and Samadpour, 2007 and references therein; Liang et al., 2015). These studies have shown that substantial populations of fecal bacteria can be often retrieved in la- goon, estuarine and coastal sediments (An et al., 2002; Luna et al., 2010; Pachepsky and Shelton, 2011; Perini et al., 2015), suggesting that sediments are environmental reservoirs of fecal bacteria. However, the reliability of these indicators has been recently questioned, as they can persist and regrow in the environment, they are recovered also in absence of obvious fecal sources, and following the discovery of environmentally-adapted populations of E. coli (Luo et al., 2011; Byappanahalli et al., 2012). Al- together, these issues have stimulated new studies, aimed at identifying alternative and more reliable indicators of fecal pollution, able to identify risks to human health and improve monitoring strategies (Stewart et al., 2008). Kreader (1995) pioneered the use of fecal anaerobes within the genus Bacteroides as more reliable fecal indi- cators, and highlighted their potential to distinguish human from non-human sources of pollution. Since then, several studies have been performed to discover and test new indicators of fecal pollution in water (reviewed in Advances in Oceanography and Limnology, 2016; 7(2): 115-124 ARTICLE DOI: 10.4081/aiol.2016.5948 This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). Next generation sequencing reveals distinct fecal pollution signatures in aquatic sediments across gradients of anthropogenic influence Gian Marco Luna,1 Grazia M. Quero,2 Laura Perini2 1CNR-ISMAR Institute of Marine Sciences, National Research Council, Largo Fiera della Pesca 2, 60125 Ancona, Italy; 2CNR-ISMAR Institute of Marine Sciences, National Research Council, Arsenale Tesa 104, 30122 Venezia, Italy *Corresponding author: gianmarco.luna@ismar.cnr.it ABSTRACT Aquatic sediments are the repository of a variety of anthropogenic pollutants, including bacteria of fecal origin, that reach the aquatic environment from a variety of sources. Although fecal bacteria can survive for long periods of time in aquatic sediments, the microbi- ological quality of sediments is almost entirely neglected when performing quality assessments of aquatic ecosystems. Here we inves- tigated the relative abundance, patterns and diversity of fecal bacterial populations in two coastal areas in the Northern Adriatic Sea (Italy): the Po river prodelta (PRP, an estuarine area receiving significant contaminant discharge from one of the largest European rivers) and the Lagoon of Venice (LV, a transitional environment impacted by a multitude of anthropogenic stressors). From both areas, several indicators of fecal and sewage contamination were determined in the sediments using Next Generation Sequencing (NGS) of 16S rDNA amplicons. At both areas, fecal contamination was high, with fecal bacteria accounting for up to 3.96% and 1.12% of the sediment bac- terial assemblages in PRP and LV, respectively. The magnitude of the fecal signature was highest in the PRP site, highlighting the major role of the Po river in spreading microbial contaminants into the adjacent coastal area. In the LV site, fecal pollution was highest in the urban area, and almost disappeared when moving to the open sea. Our analysis revealed a large number of fecal Operational Taxonomic Units (OTU, 960 and 181 in PRP and LV, respectively) and showed a different fecal signature in the two areas, suggesting a diverse contribution of human and non-human sources of contamination. These results highlight the potential of NGS techniques to gain insights into the origin and fate of different fecal bacteria populations in aquatic sediments. Key words: 16S rDNA; fecal bacteria; aquatic sediments; sewage; lagoon. Received: April 2016. Accepted: June 2016. Non -co mmerc ial us e o nly 116 Next generation sequencing of fecal bacteria McLellan and Eren, 2014), while studies of alternative fecal indicators in aquatic sediments have been rare (Kim and Wuertz, 2015). On the light of the recognized role of aquatic sediments as reservoir of fecal bacteria, the ur- gency of expanding our poor knowledge in the sedimen- tary environment becomes manifestly evident. Recent studies have shown the usefulness of Next Generation Sequencing (NGS) technologies in water qual- ity assessments (Vierheilig et al., 2015). NGS techniques can provide insights into the ecology of microbe-mediated processes (such as biodegradation of contaminants and algal blooms) influencing water quality, but can also be extremely useful to identify an array of other taxa that could serve as indicators of fecal contamination in aquatic environments (Tan et al., 2015). This holds true for those NGS methods that target small subunit rRNA hypervari- able regions, which are able to resolve microbial commu- nity composition in environmental samples, and to provide information on source-specific phylotypes and/or assemblages of phylotypes (Newton et al., 2011, 2013). Fisher et al. (2015) have recently shown the potential of sequencing the V6 region of the 16S rRNA genes to dis- cern between human versus non-human fecal sources. Cu- mulatively, these emerging studies show the potential of NGS as powerful tool in detecting alternative fecal indi- cators also in aquatic sediments. We report here the results of an investigation, carried out in two different aquatic environments in the Northern Adriatic Sea (Italy), with the aim of assessing the presence and spatial distribution of traditional and alternative (feces-associated and sewer infrastructure-associated) fecal indicators by using NGS methods targeting the 16S rRNA bacterial gene. This study is, to the best of our knowledge, among the first performed so far to explore the presence and spatial variability of alternative fecal in- dicators in different aquatic sediments, providing poten- tially useful insights for tracking the source and fate of fecal bacteria in aquatic sediments. METHODS Description of the study areas Sediments were collected in two coastal areas located in the Northern Adriatic Sea: the Po river prodelta (hereafter defined PRP) and the Lagoon of Venice (hereafter defined LV), supposed to be exposed to different types of fecal con- tamination. The PRP receives a significant discharge of contaminants from one of the largest rivers in Europe (Boldrin et al., 2005). The river discharges a mean of 1500 m3 s–1 of freshwater (with peaks up to >10,000 m3 s–1 during floods), that produces a freshwater plume able to influence the microbial diversity and functioning of the adjacent coastal ecosystem (Manini et al., 2004, Quero et al., 2015). The Po experiences typically major floods, which transport large amounts of suspended sediments and associated pol- lutants to the sea, that can be stored or transported offshore to the adjacent marine areas (Correggiari et al., 2005). Many studies have shown that the coastal area in front of the delta is severely chemically polluted, as the river carries yearly tons of anthropogenic chemicals collected from the Po valley and the river tributaries, including emerging con- taminants (Casatta et al., 2015). The LV, among the largest lagoons in the Mediterranean, is a microtidal, semi-closed lagoon connected to the Adriatic Sea by three openings (in- lets), influenced by a tidal semidiurnal regime (Cucco and Umgiesser, 2006). The tidal regime governs the water ex- change with the adjacent open sea, and largely influences the renewal capacity of the lagoon basin. The LV is histor- ically impacted by a multitude of anthropogenic stressors, among which one of the largest industrial plants of Italy lo- cated on the nearby mainland, touristic and commercial ports, several small tributaries (that carry around 30-35 m3 s–1 of freshwater; Zuliani et al., 2005), agricultural and mu- nicipal wastes. As far as the fecal contamination is con- cerned, an important source of contamination is the city of Venice which, due to its history and unique building archi- tecture, has never been provided with a modern and effi- cient sewage treatment system (Sfriso and Facca, 2013). Consequently, only partially treated effluents from a large number of domestic and commercial inputs are discharged daily into the city canals, determining a diffuse contamina- tion and the accumulation of fecal bacteria in sediments and live macroalgae (Quero et al., 2015; Perini et al., 2015). Sampling activities In the PRP area, sampling was performed between 10th and 14th June 2013 in front of the outlets of the main branches of the river delta. The sampling design included 11 stations (Fig. 1), distributed along coast-to-open sea tran- sects, at depths comprised between 9.5 and 20.5 meters. The geographic coordinates (as longitude and latitude) and water depths of the stations were as follows: 1 (12.543 E, 44.99 N, 12 m), 2 (12.556 E, 44.993 N, 20.5 m), 3 (12.571 E, 44.968 N, 9.5 m), 4 (12.576 E, 44.971 N, 15 m), 5 (12.581 E, 44.974 N, 19 m), 6 (12.573 E, 44.955 N, 10.5 m), 7 (12.585 E, 44.952 N, 14 m), 8 (12.558 E, 44.929 N, 11 m), 9 (12.573 E, 44.919 N, 14.6 m), 10 (12.538 E, 44.891 N, 14.8 m) and 11 (12.558 E, 44.886 N, 17.5 m). The sampling transects were in front of the outlets of the main branches of the river delta. The spatial distribution of the stations was thought to follow the pos- sible deposition route of pollutants transported by an ex- ceptional flood event, like the one that occurred in the third week of May 2013, with a maximum flow rate of 6830 m3 s–1. More details about the sampling activities and the characteristics of the sampling stations in the PRP are reported in Quero et al. (2015). Non -co mmerc ial us e o nly 117G.M. Luna et al. In the LV area, sampling was performed during the au- tumn season (21st-22th October 2014), chosen as the period of highest river runoff, expected to increase the load of fecal bacteria within the lagoon. The sampling design in- cluded 5 sampling stations (Fig. 1), at depths comprised between 5.4 and 16 m, distributed across a gradient of pu- tative contamination from the inner part of the lagoon to the open sea. The geographic coordinates (longitude and latitude) and water depth of the stations were as follows: Industrial Port (12.219 E, 45.438 N, 5.4 m), Inner Lagoon (12.258 E, 45.448 N, 7.2 m), Cruise Port (12.311 E, 45.436 N, 12.1 m), City Centre (12.352 E, 45.431 N, 7.1 m) and Open Sea (12.508 E, 45.313 N, 16 m). In both areas, sediments were collected using a Van Veen grab sampler (sampling surface 0.1 m2), onboard small re- search vessels (a privately operated one in the PRP area, and the ‘Litus’ boat operated by ISMAR-CNR in the LV one). Once onboard, the uppermost 0-2 cm layer of sedi- ment was immediately placed, using sterile spatulas, in sterile containers for their immediate transport at 4°C to the laboratory, where the samples were stored at -20°C until analysis. Analyses of fecal bacteria using Illumina sequencing of the 16S rRNA gene DNA was extracted from one gram of each sediment sample using the PowerSoil® DNA Isolation Kit (MoBio Fig. 1. The two study sites in the Northern Adriatic Sea (Italy), with indication and name of the sampling stations (blue dots). Latitude (N) and longitude (E) are reported; LV, Lagoon of Venice; PRP, Po river prodelta. Non -co mmerc ial us e o nly 118 Next generation sequencing of fecal bacteria Laboratories Inc., California), according to the manufac- turer’s instructions with some slight modifications to in- crease the DNA yield and quality. These modifications included two additional vortexing steps (following the one which is recommended by the manufacturer) at the max- imum speed for 2 min, each one being preceded by an in- cubation at 70°C for 5 min, and by one more washing step with Solution C5 as an additional removal step for con- taminants. The concentration of each DNA extract was determined spectrophotometrically using NanoDrop ( Thermo Scientific) and the DNA was stored at -80°C until PCR. Illumina Miseq V3 sequencing were carried out on the hypervariable V3 and V4 regions of the 16S rRNA gene by amplifying using the 341F (5′−CCTACGGGNG- GCWGCAG−3′) and 785R (5′−GACTACHVGGGTATC- TAATCC−3′) universal bacterial primers (Eiler et al., 2012). Paired-end reads were quality checked (with de- fault settings and minimum quality score of 20) and ana- lyzed with QIIME v1.8.0 software package (Quantitative Insights Into Microbial Ecology; Caporaso et al., 2010). Reads were clustered into OTUs by using UCLUST v1.2.22 (Edgar, 2010) with a >97% similarity threshold with an open-reference OTU picking strategy and default settings. Chimeras were detected by using USEARCH v6.1 (Edgar, 2010). Chimera checking and taxonomy as- signment was performed using Greengenes 13.8 as refer- ence database (DeSantis et al., 2006). To account for differences in the sequencing effort among samples, abun- dances in each sample were normalized to the number of sequences of the sample showing the lowest number of reads. The sequences of the PRP area have been submitted to the SRA (Sequence Read Archive; accession numbers SRP061637), while those for the LV are currently being submitted. Data elaboration and statistical analyses Within each of the bacterial assemblage, we searched for those OTUs identified as belonging to the traditional fecal indicator taxa (i.e., the family Enterobacteriaceae, that includes the genus Escherichia, and the genus Ente- rococcus), and those OTUs belonging to the alternative fecal indicator taxa, according to the approach proposed by Newton et al. (2013). As alternative indicators, we searched for OTUs belonging to five feces-associated bac- terial families (Bacteroidaceae, Porphyromonadaceae, Clostridiaceae, Lachnospiraceae and Ruminococcaceae) and three sewer infrastructure-associated bacterial genera (Acinetobacter, Arcobacter and Trichococcus), that we used here as signatures of fecal (human and non-human) and sewage contamination, respectively. This distinction was based on the study by Newton et al. (2013), who re- ported that five feces-associated bacterial families were prevalent (up to 85% of the total sequences) in the feces of animals and humans while, on the other hand, three sewer infrastructure-associated genera were very abun- dant in the sewage samples while not prevalent in human feces (only 33 sequences recovered out of a >1.2 million sequences of a human fecal dataset). Vandewalle et al. (2012) reported that only a small fraction of OTUs in sewage matched sequences from human fecal samples, and suggested that these sewage-associated taxa, that thrive within the sewer system, may serve as useful ad- juncts to fecal indicators for tracking sewage pollution in surface waters. The Spearman-Rank correlation analysis was performed to test for relationships between the rela- tive abundance of traditional and alternative fecal indica- tors. Correlation coefficients (r) were considered significant at P-values less than 0.05. Differences in the composition of fecal bacterial communities between the sampling areas, and between groups of stations within each of the two areas, were assessed by using the analysis of similarity (ANOSIM) tool based on a Bray-Curtis sim- ilarity matrix. The presence of statistical differences be- tween samples is indicated by a significance level at P-values less than 0.05. To assess differences in the com- position of indicator OTUs between the sampling areas, we applied univariate distance-based permutational analy- ses of variance (PERMANOVA). The statistical analyses were carried out using a sampling design, that considered the Area as fixed factor as source of variance, with 2 lev- els (PRP and LV). The ANOSIM and PERMANOVA analyses were performed using the PRIMER 6 software (http://www.primer-e.com/). RESULTS Relative abundance of traditional, feces- and sewage-associated indicators At both areas, sequence analyses of the sediment sam- ples revealed evidences for a marked and diffuse fecal sig- nature that was, however, significantly different between the two areas (ANOSIM, r=0.314, P<0.05). In the PRP area, the relative abundance of traditional indicators (Fig. 2A) accounted for 0.01 to 0.19% of the bacterial as- semblage in the case of Enterobacteriaceae, and for 0 (no sequences detected) to 0.01% in the case of Enterococcus. In the same study area, the relative abundance of alterna- tive fecal indicators was quite higher than that of tradi- tional indicators. The contribution of the feces-associated indicators (Fig. 2C) was in the range 0.08-2.13% (Lach- nospiraceae), 0.17-0.53% (Clostridiaceae), 0-0.14% (Por- phyromonadaceae), 0-0.04% (Bacteroidaceae) and 0.08-0.63% (Ruminococcaceae), while the contribution of the sewage-associated indicators (Fig. 2E) was in the range 0-0.17% (Arcobacter), 0.06-0.57% (Acinetobacter), with only one or two sequences per sample assigned to the Trichococcus genus (only at the stations 4, 5, 6 and Non -co mmerc ial us e o nly 119G.M. Luna et al. 8). The cumulative contribution of all the sequences be- longing to the traditional, feces- and sewage-associated indicators was in the range 0.49-3.96% of the bacterial as- semblages (on average 1.63%). In the LV area, the level of fecal contamination was overall reduced when compared to the PRP area. As far as the traditional indicators are concerned, these were ob- served in the LV area at all stations (Fig. 2B), in the range 0-0.05% (Enterobacteriaceae) and 0.004-0.03% (Entero- coccus). The relative abundance of the feces-associated indicators (Fig. 2D) was higher and accounted for 0.05- 0.40% (Lachnospiraceae), 0.01-0.29% (Clostridiaceae), 0-0.02% (Porphyromonadaceae), 0-0.02% (Bac- teroidaceae) and 0.001-0.27% (Ruminococcaceae), while the contribution of the sewage-associated indicators (Fig. 2F) was in the range 0.01-0.11% (Arcobacter), 0- 0.06% (Acinetobacter), with no sequences assigned to the genus Trichococcus. The cumulative contribution of all of the sequences belonging to the traditional and alterna- tive indicators accounted from 0.08 to 1.12% of the bac- terial assemblage (on average 0.60%). Spatial patterns of fecal indicators in the two areas In the PRP area, the fecal contamination in stations lo- cated closer to the coast (stations 1, 3, 4, 5, 6, 8 and 10; average relative abundance 2.0%) was two-fold higher that in offshore stations (2, 7, 9 and 11; average 0.99%). The highest abundance of fecal indicators as a whole (i.e., summing up traditional and alternative indicators) was ob- served in the stations located right in front of the main outlet of the Po river (stations 4, 5, and 6; up to 3.96% in the station 6). The relative abundance of traditional indi- cators didn’t show significant correlation either with the feces- or with the sewage-associated indicators (P≥0.05 for both relationships), whereas the feces- and sewage-as- sociated indicators were significantly and positively re- lated (r=0.83, P<0.01). In the LV area, the fecal contamination was much higher inside the lagoon (stations Ind. Port, Inner Lagoon, Cruise Port and City Centre; average relative abundance 0.73%), and only weakly detectable in the open sea station (relative abundance 0.08%). Within the lagoon, the high- est abundance of fecal indicators as a whole was observed in the station located closer to the city center of Venice (1.12%), followed by the two stations located close to the industrial area (station Ind. Port, 0.64%) and the mainland (Inner Lagoon, 0.70%). In the LV area the different types of fecal indicators were positively and significantly cor- related (r=0.82, P<0.05 between traditional and feces-as- sociated indicators; r=0.92, P<0.01 between traditional and sewage-associated indicators). The feces- and sewage-associated indicators were also significantly and positively correlated (r=0.751, P<0.05). Traditional, fecal and sewage indicator OTUs The number of OTUs that were affiliated with each of the fecal indicator groups varied widely between areas and among stations in each area, ranging from 0 to 266 Fig. 2. Bar plots of the relative abundance of 16S rDNA se- quences belonging to: A, B) the traditional fecal indicator bac- terial taxa (Enterobacteriaceae and Enterococcus); C, D) the feces-associated bacterial families (Lachnospiraceae, Clostridi- aceae, Porphyromonadaceae, Bacteroidaceae and Ruminococ- caceae); and E, F) the sewage-associated bacterial genera (Arcobacter, Acinetobacter and Trichococcus) in the sediments of the two sites. In the LV site: IP, industrial port; IL, inner la- goon; CP, cruise port; CC, city centre; OS, open sea. Non -co mmerc ial us e o nly 120 Next generation sequencing of fecal bacteria OTUs depending on the fecal indicator (Tab. 1). The re- sults of univariate PERMANOVA revealed a significant effect of the factor area (P<0.01). In the PRP area, the re- sults are summarized by grouping the stations located closer to the coast (1, 3, 4, 5, 6, 8 and 10) and those lo- cated more offshore (stations 2, 7, 9 and 11). This choice was supported by the ANOSIM analysis, that demon- strated significant differences, in terms of OTU commu- nity composition, between these two groups of stations (ANOSIM, P<0.05). In this area, the cumulative number of OTUs (i.e., the sum of all OTUs recorded at all sta- tions) belonging to the traditional indicators was 46 for Enterobacteriaceae and 4 for Enterococcus (only observed in the more coastal stations). The number of OTUs within the feces-associated indicators was much higher: 176 and 98 for Lachnospiraceae (coastal and offshore stations, re- spectively), 169 and 141 (Clostridiaceae), 55 and 36 (Por- phyromonadaceae), 22 and 9 (Bacteroidaceae), 132 and 93 (Ruminococcaceae). The number of OTUs associated with the sewage-associated indicators was 3 and 0 for Tri- chococcus (coastal and offshore stations, respectively), 69 and 54 (Acinetobacter), 19 and 34 (Arcobacter). Overall, in the PRP area the number of OTUs in the coastal stations was higher than in the offshore ones, with only one ex- ception (i.e., Arcobacter OTUs more abundant in offshore than in coastal stations). In LV area, the cumulative number of OTUs belonging to the traditional indicators was 5 and 2 (for Enterobacte- riaceae and Enterococcus, respectively). The cumulative number of the fecal-associated indicator OTUs was larger compared to the traditional ones, corresponding to 38 (Lachnospiraceae, range among sampling stations 4-18), 37 (Clostridiaceae, range 5-18), 16 (Porphyromon- adaceae, range 0-10), 18 (Bacteroidaceae, range 0-9) and 22 OTUs (Ruminococcaceae, range 0-10). At the same time, the number of OTUs belonging to the sewage-asso- ciated indicators was 12 for Acinetobacter (range 0-5) and 31 for Arcobacter (range 4-20), with no OTUs belonging to Trichococcus. The highest number of OTUs was ob- served in the lagoon stations (namely, at stations Ind. Port, Inner Lagoon and City Centre. The station located in the open sea showed only 14 OTUs for the entire pool of fecal indicators (traditional, feces- and sewage-associated). DISCUSSION In this study, we aimed at examining the presence, prevalence and spatial distribution of traditional and al- ternative fecal indicators in a range of aquatic sediments, collected in transitional, estuarine and coastal marine areas along gradients of anthropogenic influence. We used NGS of 16S rRNA gene amplicons to identify and track fecal indicator bacteria within complex benthic microbial assemblages, by taking advantage of the method recently proposed by Newton et al. (2013), that allows to poten- tially discriminate between feces-associated and sewage- associated bacteria. The presence and distribution patterns of alternative indicators in the sediments were then com- pared with those of traditional indicators (Escherichia coli, within the Enterobacteriaceae family, and entero- cocci) that are utilized worldwide to assess fecal pollution Tab. 1. Number of OTUs affiliated with the different fecal indicator bacteria in the two study sites. For the Po River prodelta site, results are summarized by summing all the OTUs observed at stations 1, 3, 4, 5, 6, 8 and 10 (stations close to the coast) and those observed at stations 2, 7, 9 and 11 (offshore stations). PRP LV Microbial Number of OTUs Number of OTUs Cumulative indicator Stations close Offshore Cumulative Stations within the lagoon no. of OTUs to the coast stations no. of OTUs IP IL CP CC Open sea Traditional Enterobacteriaceae 27 28 46 1 5 1 3 0 5 Enterococcus 4 0 4 1 1 1 3 1 2 Feces-associated Lachnospiraceae 176 98 266 17 18 9 9 4 38 Clostridiaceae 169 141 231 10 16 18 10 5 37 Porphyromonadaceae 55 36 70 3 7 2 10 0 16 Bacteroidaceae 22 9 28 3 9 4 8 0 18 Ruminococcaceae 132 93 187 6 10 10 6 0 22 Sewage-associated Trichococcus 3 0 3 0 0 0 0 0 0 Acinetobacter 69 54 85 5 4 1 3 0 12 Arcobacter 19 34 40 20 9 8 17 4 31 PRP, Po River prodelta; LV, lagoon of Venice; OTU, Operational Taxonomic Units; IP, industrial port; IL, inner lagoon; CP, cruise port; CC, city centre; OS, open sea. Non -co mmerc ial us e o nly 121G.M. Luna et al. in aquatic environments. Recently, a large body of studies have identified and tested alternative indicators of fecal pollution in aquatic systems, by exploiting the NGS tech- nologies to provide unprecedented inventories of micro- bial communities in aquatic samples (McLellan and Eren, 2014). However, while a large number of studies have in- vestigated the presence and spatial patterns of alternative indicators in water (Savichtcheva and Okabe, 2006; Liang et al., 2015), similar studies in sediments have been rare (Kim and Wuertz, 2015). This lack of knowledge is sur- prising, given the recognized role of aquatic sediments as reservoir of fecal bacteria (Luna et al., 2010), and their potential to favor the persistence of fecal microbes and to contaminate the overlying water through resuspension, which likely poses important public health and environ- mental threats. At both study areas, we found evidences for an exten- sive microbial pollution that was testified, despite at dif- ferent extent, by the presence of traditional, feces- and sewage-associated indicators. These indicators accounted cumulatively for a variable fraction of benthic assem- blages with peaks, as in the case of the most polluted sed- iments, up to 3.96%. At both study areas, the magnitude of the fecal signature decreased with increasing distance from the sources of pollution, which underlines the use- fulness of this approach, that includes also the alternative indicators, to track the fecal pollution in aquatic sedi- ments. The presence of fecal pollution in the two sites was not unexpected, since they are both subjected to a signif- icant anthropogenic pressure, and likely receive important loads of fecal bacteria. The PRP coastal area receives a significant discharge of sediments and pollutants from the largest Italian river, especially after the floods that occur on a seasonal (biannual) frequency and, also, as episodic short-term events (Palinkas et al., 2005). Its drainage basin contributes to more than 35% of the national agri- cultural, livestock and industrial production, that origi- nates organic loads estimated in 114×106 inhabitant equivalents (Casatta et al., 2015). However, while the presence of chemical pollutants in the coastal area near the mouth of the Po river has been largely reported (Ro- mano et al., 2013), this is the first report on the presence of extensive fecal pollution in these sediments. The largest fecal contamination was observed in the stations located in front of mouth of the Pila distributary, which is the main outlet of the river and can discharge up to 70% of the sediment load delivered to the sea. However, the fecal signature was also observed in stations located down- stream of this main mouth, that suggests either a fecal input from the other minor distributaries, and/or the po- tential of fecal bacteria of being distributed over a large coastal area. This bacterial spread may occur when the high fluvial discharge coincides with energetic physical oceanographic conditions, preventing deposition in shal- low waters and favoring sediment dispersion in seabed deposits located offshore and downstream of the mouth (Palinkas et al., 2005). Our finding of a large reservoir of fecal bacteria in the Po prodelta sediments poses a claim for potential health consequences, given the potential of sediment bacteria of being re-suspended and transported southward, where several bathing beaches and touristic destinations are present. Nevertheless we point out here that further studies are needed to investigate the ability of the different fecal bacterial populations, including the feces- and sewage-associated bacteria, to persist or decay once they reach the marine environment, and of being po- tentially transported toward adjacent coastal areas in pres- ence of specific hydrological conditions. The finding of a diffuse fecal contamination in the LV area, especially in the stations more exposed to anthro- pogenic impacts, likely depends on the wide variety of human impacts, in the form of large industrial plants, touris- tic and commercial ports, rivers, agricultural and municipal wastes, that have affected this vulnerable transitional envi- ronment in the last years (Micheletti et al., 2011). Our re- sults show that the highest fecal contamination is observed in the area closer to the city of Venice, confirming recent findings of a chronic fecal pollution (Perini et al., 2015), but also that other areas of the Venice lagoon, located closer to the mainland, receive important loads of bacteria of fecal origin. This may be due to multiple delivery routes dis- charging in the inner part of the lagoon, that include runoff, outfall discharge, sewer overflows (that causes untreated sewage to be released), presence of tributaries, and other human activities. It is worth noticing that a signal, despite weak, of fecal contamination was observed also in the sed- iments of the open sea station. We speculate that the pres- ence of fecal pollutants in this area, which is relatively far by obvious sources of pollution, could be the consequence of fecal discharges by offshore point sources and/or by the underwater submarine wastewater pipes present in the area (Scroccaro et al., 2010), that may disperse microbial pol- lutants over vast marine areas. We report here that different aquatic sediments sub- jected to anthropogenic influence contain important pro- portions of traditional indicator bacteria, as well as fecal bacteria representing signatures of fecal (human and ani- mal) and sewage sources. At both areas, the relative abun- dance of fecal- and sewage-associated bacteria always exceeded that of the traditional indicators. Sediments are a potentially favorable environment for fecal microorgan- isms survival, and this may be especially true for many of the alternative indicators (such as Bacteroidales), that do not survive well in water due to their obligate anaerobic physiology (Bae and Wuertz, 2015) but may persist longer in low-oxygen or anoxic layers that are common in sedi- ments. In both areas, we found that the fecal signature was more evident than the sewage signature. However, the av- Non -co mmerc ial us e o nly 122 Next generation sequencing of fecal bacteria erage ratio of the fecal pollution signature to the sewage pollution signature was different in the two areas. In PRP, the average contribution of the feces-associated taxa was 1.244%, a percentage 3.9-fold higher than that of the sewage-associated taxa (0.320%). Conversely, in the LV area, the average contribution of feces-associated taxa was 0.602% (excluding the open sea station), a percentage 7.05-fold higher than that of the sewage-associated indi- cators (0.085%). The prevalence of different fecal bacterial signatures in the two areas suggests the presence of differ- ent fecal sources. The higher ratio of feces- to sewage-as- sociated bacteria observed in the lagoon of Venice suggests that human/animal sources of pollution are more important than sewage in polluting this area while, in the Po prodelta area, the sewage pollution appears to be an important source. The sewage pollution in the PRP area is particu- larly evident in the Acinetobacter signature which is, on average, 14-fold more abundant than in the LV site. This genus is the most important in sewage (Newton et al., 2013), and its recovery in sediments suggests that it may be a reliable signature of sewage pollution also in the ben- thic environment. It remains to be determined whether members of this genus are adapted to survive in aquatic sediments, and how long are their decay rates once they reach the sedimentary environment. The importance of sewage in contributing to pollution in the Po prodelta is also confirmed by the recovery of another sewage-associ- ated genus (Trichococcus) that, though with only a few se- quences per sample, was observed only in this area and not in the Venice lagoon. A very large number of fecal OTUs, accounting for a total of 960 and 181 OTUs (in the PRP and LV areas, re- spectively), were observed in the sediments under scrutiny. The largest number of OTUs was observed in the Po prodelta area. This may suggest the existence, in this area, of multiple delivery mechanisms containing multiple fecal sources, that likely originate from the large drainage basin, and resulting in more complex fecal OTU signa- tures than those observed in the LV sediments. In many cases and at both areas, fecal bacteria populations were dominated by few dominant OTUs. A closer examination of the top most abundant fecal OTUs in the two sites pro- vided additional insights, useful to potentially distinguish among different sources of pollution. In the PRP area, the traditional indicators showed dominance of one Enter- obacteriaceae OTU (range 5-177 sequences per sediment sample) that showed 100% Blast identity with Es- cherichia coli strain 732 (accession number CP015138). The enterococci populations, that were poorly represented at this area, were dominated by one OTU affiliated with Enterococcus casseliflavus isolated from cow rumen (ac- cession number KT630829), suggesting that cattle may be an important route of fecal pollution in this area. Con- versely, in the LV area, the dominant Enterobacteriaceae OTU was affiliated with Yersinia kristensenii (accession number HG938308.1) while the dominant enterococcal OTU, especially abundant in the station closer to the city center, showed a top Blast match with Catellicoccus mari- mammalium (accession number KF251005.1) isolated from gulls’ feces. In the Venice urban area, populations of urban gulls have increased exponentially in the last years (Rock, 2012). Our findings indicate that they can contribute as an additional source of fecal contamination in the lagoon, which deserves further investigations. As far as the fecal- and sewage-associated OTUs are concerned, the same analysis of the identity of the domi- nant OTUs revealed also interesting additional insights. The dominant Acinetobacter OTUs in the LV area, par- ticularly abundant in the station closer to the city center, showed a top Blast match with an uncultured Acinetobac- ter (HQ742373.1) associated with the human intestine, confirming human fecal pollution as an important con- tamination route in the city (Perini et al., 2015). Con- versely, the analyses of some of the dominant Lachnospiraceae OTUs showed, at both PRP and LV areas, top Blast identities with several uncultured bacteria, which were observed in a variety of sediments and also in coastal sediments vegetated by seagrasses (Jensen et al., 2007), apparently far from fecal pollution sources. These preliminary findings suggest that some of members within this family may be part of the natural benthic as- semblages, and are thus not reliable indicators of fecal pollution. It is evident that, given also the much higher complexity of benthic microbial assemblages when com- pared with the planktonic ones, there is still much to be deciphered when using this type of community ap- proaches to track fecal pollution in aquatic sediments. Overall, our results demonstrate that the coupled analyses of the diversity and magnitude of the three fecal signa- tures (traditional, fecal- and sewage-associated bacteria) are useful to discern among different pollution sources in the sediments of transitional, estuarine and coastal areas. CONCLUSIONS We have shown that a wide range of aquatic sediments in areas exposed to anthropogenic stressors host important proportions of traditional indicator bacteria, but also of alternative bacterial taxa that more specifically track the presence of fecal (human and animal) and sewage pollu- tion. The magnitude and pattern of the complex fecal sig- nature followed the expected gradients of microbial pollution, and provided information useful to identify the main sources of pollution in the two study sites. Our re- sults also emphasize the opportunities that NGS tech- niques now offer to disentangle complex fecal pollution signals, and to source track and identify alternative fecal indicators in lagoon and marine sediments. Non -co mmerc ial us e o nly 123G.M. Luna et al. ACKNOWLEDGMENTS This work was possible thanks to funds granted to GML by the Programme RITMARE (SP3-WP2-A2 Stru- menti innovativi per la valutazione degli effetti di conta- minanti emergenti sulle comunità biologiche), the IPA Project BALMAS - Ballast Water Management System for Adriatic Sea Protection (project code 1° STR/0005) funded by EU, the Short Term Mobility (STM) Program funded by the CNR, and partially to funds granted to GMQ by the Malware - Metagenomic study of microbial invasive species introduced by ballast waters project (funded by Lifewatch Italia). We sincerely thank Dr. Francesca Garaventa (ISMAR-CNR) for her help in sam- pling the PRP site, and similarly the crew of the research boat ‘Litus’ and Dr. Mauro Bastianini (ISMAR-CNR) for making sample collection possible in the LV site. REFERENCES An YJ, Kampbell DH, Breidenbach GP, 2002. Escherichia coli and total coliforms in water and sediments at lake marinas. Environ. Pollut. 120:771-778. Bae S, Wuertz S, 2015. Decay of host-associated Bacteroidales cells and DNA in continuous-flow freshwater and seawater microcosms of identical experimental design and tempera- ture as measured by PMA-qPCR and qPCR. Wat. Res. 70: 205-213. Byappanahalli MN, Nevers MB, Korajkic A, Staley ZR, Har- wood VJ, 2012. Enterococci in the environment. Microbiol. Mol. Biol. Rev. 76:685-706. Boldrin A, Langone L, Miserocchi S, Turchetto M, Acri F, 2005. Po River plume on the Adriatic continental shelf: dispersion and sedimentation of dissolved and suspended matter during different river discharge rates. Mar. Geol. 222:135-158. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Fierer N, Pena AG, Goodrich JK, Gordon JI, Huttley GA, Kelley ST, Knights D, Koenig JE, Ley RE, Lozupone CA, McDonald D, Muegge BD, Pirrung M, Reeder J, Sevinsky JR, Turnbaugh PJ, Walters WA, Wid- mann J, Yatsunenko T, Zaneveld J, Knight R, 2010. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7:335-336. Casatta N, Stefani F, Pozzoni F, Guzzella L, Marziali L, Mascolo G, Viganò L, 2015. Endocrine-disrupting chemicals in coastal lagoons of the Po River delta: sediment contamina- tion, bioaccumulation and effects on Manila clams. Environ. Sci. Pollut. Res. Int. 23:10477-10493. Correggiari A, Cattaneo A, Trincardi F, 2005. The modern Po Delta system: lobe switching and asymmetric prodelta growth. Mar. Geol. 222:49-74. Cucco A, Umgiesser G, 2006. Modeling the Venice Lagoon res- idence time. Ecol. Model. 193:34-51. DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, Keller K, et al., 2006. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72:5069-5072. Edgar RC, 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26:2460-2461. Eiler A, Heinrich F, Bertilsson S, 2012. Coherent dynamics and association networks among lake bacterioplankton taxa. ISME J. 6:330-342. Elliott JE, Elliott KH, 2013. Tracking marine pollution. Science 340: 556-558. Field KG, Samadpour M, 2007. Fecal source tracking, the indi- cator paradigm, and managing water quality. Water Res. 41:3517-3538. Fisher JC, Eren AM, Green HC, Shanks OC, Morrison HG, Vineis JH, Sogin ML, McLellan SL, 2015. Comparison of sewage and animal fecal microbiomes by using oligotyping reveals potential human fecal indicators in multiple taxo- nomic groups. Appl. Environ. Microbiol. 81:7023-7033. Halpern BS, Walbridge S, Selkoe KA, Kappel CV, Micheli F, D’Agrosa C, et al., 2008. A global map of human impact on marine ecosystems. Science 319:948-952. Jensen SI, Kühl M, Priemé A, 2007. Different bacterial commu- nities associated with the roots and bulk sediment of the sea- grass Zostera marina. FEMS Microbiol. Ecol. 62:108-117. Kim M, Wuertz S, 2015. Survival and persistence of host-associ- ated Bacteroidales cells and DNA in comparison with Es- cherichia coli and Enterococcus in freshwater sediments as quantified by PMA-qPCR and qPCR. Water Res. 87:182-192. Kreader CA, 1995. Design and evaluation of Bacteroides DNA probes for the specific detection of human fecal pollution. Appl. Environ. Microbiol. 61: 1171-1179. Liang L, Goh SG, Vergara GGRV, Fang HM, Rezaeinejad S, Chang SY, Bayen S, Lee WA, Sobsey MD, Rose JB, Gin KYH, 2015. Alternative fecal indicators and their empirical relationships with enteric viruses, Salmonella enterica, and Pseudomonas aeruginosa in surface waters of a tropical urban catchment. Appl. Environ. Microbiol. 81:850-860. Luna GM, Vignaroli C, Rinaldi C, Pusceddu A, Nicoletti L, Gabellini M, Danovaro R, Biavasco F, 2010. Extraintestinal Escherichia coli carrying virulence genes in coastal marine sediments. Appl. Environ. Microbiol. 76:5659-5668. Luo C, Walk ST, Gordon DM, Feldgarden M, Tiedje JM, Kon- stantinidis KT, 2011. Genome sequencing of environmental Escherichia coli expands understanding of the ecology and speciation of the model bacterial species. P. Natl. Acad. Sci. USA 108:7200-7205. Manini E, Luna GM, Danovaro R, 2004. Benthic bacterial re- sponse to variable estuarine water inputs. FEMS Microbiol. Ecol. 50:185-194. McLellan SL, Eren AM, 2014. Discovering new indicators of fecal pollution. Trends Microbiol. 22:697-706. Micheletti S, Gottardo A, Critto A, Chiarato A, Marcomini A, 2011. Environmental quality of transitional waters: the la- goon of Venice case study. Environ. Int. 37:31-41. Newton RJ, Bootsma MJ, Morrison HG, Sogin ML, McLellan SL, 2013. A microbial signature approach to identify fecal pollution in the waters off an urbanized coast of Lake Michi- gan. Microb. Ecol. 65:1011-1023. Newton RJ, VandeWalle JL, Borchardt MA, Gorelick MH, McLellan SL, 2011. Lachnospiraceae and Bacteroidales al- ternative fecal indicators reveal chronic human sewage con- tamination in an urban harbor. Appl. Environ. Microbiol. 77: 6972-6981 Non -co mmerc ial us e o nly 124 Next generation sequencing of fecal bacteria Pachepsky YA, Shelton DR, 2011. Escherichia coli and fecal coliforms in freshwater and estuarine sediments. Crit. Rev. Environ. Sci. Technol. 41: 067-1110. Palinkas CM, Nittrouer CA, Wheatcroft RA, Langone L, 2005. The use of 7 Be to identify event and seasonal sedimentation near the Po River delta, Adriatic Sea. Mar. Geol. 222:95-112. Perini L, Quero GM, García ES, Luna GM, 2015. Distribution of Escherichia coli in a coastal lagoon (Venice, Italy): tem- poral patterns, genetic diversity and the role of tidal forcing. Water Res. 87:155-165. Quero GM, Cassin D, Botter M, Perini L, Luna GM, 2015. Pat- terns of benthic bacterial diversity in coastal areas contam- inated by heavy metals, polycyclic aromatic hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs). Front. Mi- crobiol. 6:1053. Ridgway J, Shimmield G, 2002. Estuaries as repositories of his- torical contamination and their impact on shelf seas. Est. Coast. Shelf Sci. 55: 903-928. Rock P, 2012. Urban gulls. Why current control methods always fail. Riv. Ital. Ornitol. 82: 58-65. Romano S, Langone L, Frignani M, Albertazzi S, Focaccia P, Bellucci LG, Ravaioli M, 2013. Historical pattern and mass balance of trace metals in sediments of the northwestern Adriatic Sea Shelf. Mar. Poll. Bull. 76: 32-41. Savichtcheva O, Okabe S, 2006. Alternative indicators of fecal pollution: relations with pathogens and conventional indi- cators, current methodologies for direct pathogen monitor- ing and future application perspectives. Water Res. 40: 2463-2476. Scroccaro I, Ostoich M, Umgiesser G, De Pascalis F, Colugnati L, Mattassi G, Vazzoler M, Cuomo M, 2010. Submarine wastewater discharges: dispersion modelling in the Northern Adriatic Sea. Environ. Sci. Pollut. Res. 17: 844-855. Sfriso A, Facca C, 2013. Annual growth and environmental re- lationships of the invasive species Sargassum muticum and Undaria pinnatifida in the lagoon of Venice. Est. Coast. Shelf Sci. 129:162-172. Stewart JR, Gast RJ, Fujioka RS, Solo-Gabriele HM, Meschke JS, Amaral-Zettler LA, del Castillo E, Polz MF, Collier TK, Strom MS, Sinigalliano CD, Moeller PDR, 2008. The coastal environment and human health: microbial indicators, pathogens, sentinels and reservoirs. Environ. Health 7:S3. Tan B, Ng C, Nshimyimana JP, Loh LL, Gin KY-H, Thompson JR, 2015. Next-generation sequencing (NGS) for assessment of microbial water quality: current progress, challenges, and future opportunities. Front. Microbiol. 6:1027. Vandewalle JL, Goetz GW, Huse SM, Morrison HG, Sogin ML, Hoffmann RG, Yan K, McLellan SL, 2012. Acinetobacter, Aeromonas and Trichococcus populations dominate the mi- crobial community within urban sewer infrastructure. Env- iron. Microbiol. 14: 2538-2552. Vierheilig J, Savio D, Ley RE, Mach RL, Farnleitner AH, Reis- cher GH, 2015. Potential applications of next generation DNA sequencing of 16S rRNA gene amplicons in microbial water quality monitoring. Wat Sci Technol 72:1962-1972. Vörösmarty CJ, McIntyre PB, Gessner MO, Dudgeon D, Pru- sevich A, Green P, Glidden S, Bunn SE, Sullivan CA, Lier- mann CR, Davies PM, 2010. Global threats to human water security and river biodiversity. Nature 67:555-561. Zuliani A, Zaggia L, Collavini F, Zonta R, 2005. Freshwater dis- charge from the drainage basin to the Venice Lagoon (Italy). Environ. Internat. 31:929-938. Non -co mmerc ial us e o nly