TY - JOUR
AU - Field, Erin, K
AB - ABSTRACT Microorganisms attached to aquatic steel structures play key roles in nutrient cycling and structural degradation processes. Corrosion-causing microbes are often the focus of studies involving microbially influenced corrosion, yet the roles of remaining community members remain unclear. This study characterizes the composition and functional potential of a ‘core steel microbiome’ across stainless steel types (304 and 316) and historic shipwreck steel along salinity gradients in North Carolina estuaries. We found higher phylogenetic evenness and diversity on steel surfaces compared to sediment, and at lower salinities. The core steel microbiome was composed of heterotrophic generalist taxa, and community composition was most strongly influenced by salinity. Substrate type was a secondary factor becoming more influential at higher salinities. The core steel microbiome included members of Sphingobacteriia, Cytophagia, Anaerolineaceae, Verrucomicrobiaceae, Chitinophagaceae, and Rheinheimera. While salinity differences led to phylogenetic separations across microbial community assemblages, functional genes were conserved across salinity and steel type. Generalist taxa on steel surfaces likely provide functional stability and biofilm protection for the community with limited functional trade-offs compared to surrounding environments. Further, characterization of a core steel microbiome increases the understanding of these complex steel surface microbial communities and their similarities to core microbiomes in other environments. microbial surface colonization, stainless steel, shipwrecks, biofilms, biocorrosion, core microbiome INTRODUCTION Microbially influenced corrosion (MIC) has been well described for its corrosion-causing community members and significant global costs of infrastructure destruction (Beech and Sunner 2004; Little and Lee 2007; Emerson, Fleming and McBeth 2010; Dang and Lovell 2016; Emerson 2018). Much less is known about the core microbial community members ‘behind the scenes’ which promote biofilm formation and functional stability among these MIC sites on steel surfaces. Recent evidence suggests that the abundance of corrosion-causing community members varies based on environmental factors such as substrate, salinity, and successional timing (McBeth et al. 2011; McBeth, Fleming and Emerson 2013; McBeth and Emerson 2016; Garrison, Price and Field 2019; Price et al. 2020). If the core microbial members are also shifting in community composition, then the potential for biocorrosion and nutrient-cycling ecosystem services (i.e. carbon, nitrogen and phosphorous) are likely also impacted. Characterization of the entire microbial community is important, because less dominant and rare microbial community members maintain important roles in shaping community function (Hol et al. 2010; Shade et al. 2014; Aanderud et al. 2015; Kleindienst et al. 2016; Jousset et al. 2017). The characterization of a ‘core steel microbiome’ across variations in environmental conditions could greatly improve the understanding of steel surface community functional potential. The distribution and abundance of functional genes within these steel microbial communities are also largely unknown and could provide further supporting evidence of the core steel microbiome functional role across different environmental conditions, as well as the overall functional potential of these microbial communities colonizing steel surfaces. Microbial communities on both mild steel and stainless steel are highly diverse (Vandecandelaere et al. 2010; McBeth et al. 2011), and thus the presence of many non-corroding organisms must also help provide favorable conditions for biocorrosion, nutrient cycling, and overall community function. Aquatic sediment microbial communities often have higher phylogenetic diversity and higher alpha diversity than those found in the water column (Lozupone and Knight 2007; Feng et al. 2009), but data is limited for diversity indices of steel surfaces in aquatic environments. Phylogenetic species evenness has been shown to have a positive correlation with bacterial productivity and overall functional stability (Wittebolle et al. 2009; Haig et al. 2015; Schmidt et al. 2020), and can also be used as an indicator for core microbiomes (Helmus et al. 2007; Shade and Handelsman 2012). Comparisons of community diversity indices between steel surfaces and other substrates (i.e. sediment and surface controls) could broaden the overall understanding of steel surface microbial function and a potential core steel microbiome. The characterization of a core steel microbiome could be an important step in determining the conditions under which steel is most vulnerable to microbially influenced corrosion, or even conversely, more likely to exhibit corrosion inhibition. This is further strengthened by the fact that biocorrosion is a complex process that has yet to be definitively connected to a single biochemical reaction or a single group of microbes, and with many different protein secretions and metabolites likely playing important secondary roles in biocorrosion (Kip and van Veen 2015). The importance of characterizing a core microbiome has recently expanded beyond host associated environments in order to more accurately define the health of microbial communities and their potential response to ecological disturbances (Shade and Handelsman 2012). If there are common microbial members across multiple assemblages within the same habitat after sampling across space and time, then those members may comprise a core microbiome (Berg et al. 2020). These core members may be governed by either functional or phylogenetic redundancy, and determining the degree of each could indicate how well these communities are adapted to and/or prepared for global change (Shade and Handelsman 2012). The dynamic environmental conditions within estuaries and the wide variety of steel types available in these environments for colonization could be analogous to global change. With the use of omics techniques to identify both phylogeny and function, this study aims to characterize a core steel microbiome based on the similarities of steel microbial communities across variables including salinity, steel type, and timescale, while also highlighting the differences in functional potential between communities on steel and control substrates. The influence of salinity on overall microbial communities is well known for aquatic systems and is the most significant environmental factor regulating phylogenetic divergence and community composition (Bouvier and del Giorgio 2002; Lozupone and Knight 2007; Herlemann et al. 2011). However, the degree to which salinity controls bacterial diversity is not always consistent. Higher salinity environments have been shown to have higher phylogenetic diversity, i.e. decreased taxonomic relatedness (Lozupone and Knight 2007), while other studies have shown decreasing bacterial richness, evenness, and diversity with increasing salinity (Oren 2002; Wang et al. 2012; Nawar 2016). Others have even shown no clear trend between salinity and bacterial alpha diversity (Hu et al. 2016). Uncovering the bacterial diversity relationship for steel surface environments will indicate whether phylogenetic divergence occurs in core steel microbiomes across salinity, as well as provide insight into potential differences in bacterial productivity and functional stability. Timescale is also important to consider, as corrosion-causing organism abundance can shift from iron-oxidizer dominated to sulfate-reducer dominated in a matter of days (McBeth and Emerson 2016). If the core steel community is largely the same or similar across time, then any changes in corrosion-causing organism niche establishment could be more appropriately inferred as a direct cause of chemical or abiotic factors rather than shifts in the core steel microbiome. The chemical composition of steel substrates often changes over long timescales due to oxidation and reduction microbial processes, which create metabolic byproducts such as hydrogen sulfides, iron oxides, and carbon-based organics (Emerson 2018) which could presumably also influence core microbial member assembly. While it may be difficult to say how these communities shift over timescales of years without extremely long-term experiments, historical ferrous-hulled shipwrecks provide a mechanism for studying in situ long-term effects. A study involving the same shipwreck used in this study showed that relative iron-oxidizer abundance varied significantly based on sampling location on the ship (Price et al. 2020). Further analysis of the whole microbial community from these samples could help determine whether corrosion-causing organism abundance is more heavily controlled by abiotic or biotic factors. Combining data from these shipwreck samples with short-term stainless steel coupon deployment samples (Garrison, Price and Field 2019) allows a core steel microbiome to be characterized across salinity gradients, three different steel types, and two different timescales. The increased combinations of environmental conditions found in this study provide a more encompassing view of the microbial communities associated with steel and the role of their environment. An analysis of microbial community composition and function was conducted using 16S rRNA gene sequence data and metagenomic sequencing data from three different types of steel surfaces in North Carolina estuarine environments: 304 stainless steel (304SS) (18% chromium and 8% nickel), 316 stainless steel (316SS) (16.6% chromium, 10% nickel, and 2% molybdenum), and historical ferrous-hulled shipwreck steel (exact composition unknown). These previous studies found that the abundance of corrosion-causing microbes colonizing the steel samples varied with environmental factors (Garrison, Price and Field 2019; Price et al. 2020), suggesting that community functional genes and the presence of a core microbiome may also vary with these factors and have important impacts on biocorrosion potential. This study enables a wholistic view of aquatic steel microbial communities that builds upon these previous foundational studies. Microbial community phylogeny and functional genes were also analyzed within nearby sediment samples and surface control samples (PVC biofoul plates and empty oyster shells attached to the shipwreck) to further illustrate the uniqueness of steel surfaces as a novel ecological niche. These comparisons further allow us to determine how steel surfaces may have selected for specific microbial assemblages and how core metabolic functions have adapted to these unique environments. In contrast to previous MIC community studies which often lack metagenomic sequencing, this study provides some context for functional genes found on steel surfaces in addition to characterizing a core steel microbiome. METHODS Field deployments and sample processing Stainless steel coupons were collected in February and May 2017 after being deployed for 57 and 40 days respectively, from five sites on the Pamlico River and five sites on the Neuse River, North Carolina, with salinities ranging from 0.1 ppt to 15.8 ppt (Garrison, Price and Field 2019). Coupons were deployed along two estuarine river systems in NC and processed according to the methods outlined in Garrison, Price and Field (2019). Substrate types for analysis from these stainless steel sites included 304SS, 316SS, nearby sediment, and biofouling plates made out of polyvinyl chloride (PVC) plastic as a non-metal control. Plastics such as PVC are also susceptible to fouling and deterioration but do not undergo corrosion in terms of iron oxidation. Stainless steel and PVC have no difference in aquatic bacterial colonization potential (Pedersen 1990), which makes them ideal surface controls to compare to steel biocorrosion communities (Garrison, Price and Field 2019; Price et al. 2020). Shipwreck samples were collected in September 2017 from a ferrous-hulled World War II gunboat located in the Pamlico Sound, North Carolina, at a salinity of 17.9 ppt (Price et al. 2020). Shipwreck sample types for analysis included metal pieces of the ship that appeared visibly corroded (either orange, black, or both orange and black), or not visibly corroded. All non-visibly corroded ship pieces consisted of a mixture of empty oyster shells attached to ship substrate (referred to herein as oyster surface controls). Nearby sediment was also analyzed for community comparisons to steel surfaces, according to the methods outlined in Price et al. (2020). Sampling site details are listed in Table S1 (Supporting Information). 16S rRNA sequencing and analysis This study analyzed 16S rRNA gene amplicon sequencing data to determine microbial community composition from samples related to stainless steel sites (Garrison, Price and Field 2019) and shallow water shipwreck sites (Price et al. 2020) (Table S1, Supporting Information). Extracted DNA were sent to CGEB Integrated Microbiome Resource (Dalhousie University, Halifax, Canada) for sequencing. 16S sequences were processed using a Phusion polymerase PCR amplification method (Comeau, Douglas and Langille 2017), targeting the V4-V5 region of the 16S rRNA gene, and sequenced using the Illumina MiSeq platform. The V4-V5 region is the most reliable hypervariable region for representing the full length of the 16S rRNA gene and the largest majority of bacteria phylogeny (Yang, Wang and Qian 2016). The V6-V8 region is often able to capture a larger diversity of marine bacterial taxa (Willis, Desai and LaRoche 2019); however, the sampling sites in this study are strongly influenced by terrestrial and freshwater environments, which is why the v4-v5 region was used. Amplicon sequence data was processed and annotated using the mothur v1.41.3 pipeline (Schloss et al. 2009). Output from mothur based on 97% OTU classifications was used to create CCA and NMDS ordination plots using a Bray–Curtis similarity matrix method in R with vegan and ggplot2 packages and the metaMDS function (R Core Team 2013). Significant differences between groups in CCA and NMDS plots were tested using the ANOSIM function in R. Adonis function was used to test significance and degree of correlation of environmental variables within the CCA community composition plots. SIMPER function was used to determine the most influential taxa towards differences seen in microbial community compositions between sample types. Community diversity indices were calculated using the diversity function within vegan. Linear regression of Shannon diversity was tested using the glm function within the tidyverse package. Significant differences between diversity index values were tested using the wilcox.test function in R. An indicator species analysis was performed using the indicspecies package in R to determine OTUs within the dataset that were deemed indicators for certain sample types and/or environmental conditions such as steel surface associated samples (De Cáceres and Legendre 2009). The indicspecies function determined probability percentages based on exclusivity (whether or not the OTU occurred only in that sample type) and fidelity (whether or not the OTU occurred at all sites within that sample type). Levins’ niche breadth analysis was further used to determine whether OTUs within the dataset could be categorized as generalists or specialists, and thus whether or not certain sample types contained higher proportions of either category. Niche breadth values were calculated using the formula |$B\ = \ 1/{{\Sigma }}_i$|pi2, where |$B$| is niche breadth and pi is proportion of OTUs within sample i (Levins 1968). Three-dimensional surface plots were created using Matlab v.R2017b to visualize distributions of specialists and generalists across sample types. Shotgun metagenomic sequencing and analysis Shotgun metagenomic sequencing was used to determine the functional genes present on eight samples from two of the stainless steel sites with salinities of 3.6 ppt and 6 ppt (Site N3 and N11; Table S1, Supporting Information). Samples included 304SS, 316SS, sediment, and PVC biofoul surface control for each site. Both 316SS samples from these sites were orange and visibly corroded. Eight samples from the shipwreck site with salinity of 17.9 ppt were also sequenced for shotgun metagenomics. Samples included three visibly corroded ship pieces, one sediment sample, one oyster surface control, one surrounding seawater sample, and two drilled shipcore samples extracted from the interior of the ship's hull. One 304SS sample was not included in the metagenomic results due to a low number of raw reads resulting in few genes recovered compared to the other samples. Metagenomic samples were prepared using an Illumina Nextera Flex kit at CGEB Integrated Microbiome Resource center (Comeau, Douglas and Langille 2017). Samples were enzymatically sheared and tagged with adaptors, PCR amplified using barcodes, then purified, normalized, and pooled for loading. Samples were sequenced using the Illumina NextSeq platform, with 2X sequencing depth for ∼8 million paired-end 150+150 bp reads (16 million single reads) resulting in ∼2.4 Gb per sample. Metagenomic raw reads were trimmed and confirmed using TrimGalore v.0.4.5 (Krueger 2012) and FastQC v.0.11.8 (Andrews 2010), respectively. Trimmed reads were concatenated, then assembled using SPAdes v.3.13.0 (Bankevich et al. 2012). Assembled contigs were functionally annotated using MG-RAST v.4.0.3 (Meyer et al. 2008) to search for functional genes of interest. Significant differences between metagenomic gene abundance were tested using the wilcox.test function in R. FeGenie (Garber et al. 2020), LithoGenie (https://github.com/Arkadiy-Garber/LithoGenie), Prodigal v2.6.3 (Hyatt et al. 2010), and HMMer v.3.3 (https://hmmer.org/) were used for gene annotations and estimating MIC-related protein abundance. Assembled metagenomic contigs were further binned into metagenome assembled genomes (MAGs) using MaxBin v.2.2.7 (Wu, Simmons and Singer 2016), Concoct v.1.1.0 (Alneberg et al. 2014), and MetaBAT v.2.12.1 (Kang et al. 2019). Additional program dependencies utilized by these programs included bowtie v.2.3.5.1 (Langmead and Salzberg 2012) and samtools v.1.9 (Li et al. 2009). Results from all three binning methods were analyzed using DAS Tool v.1.1.2 (Sieber et al. 2018) to determine the MAGs that had the highest completeness (>50%) and lowest contamination (<10%). Quality and bin statistics were further confirmed using CheckM v.1.0.18 (Parks et al. 2015), which also relies on pplacer v.1.1.alpha19 (Matsen, Kodner and Armbrust 2010), prodigal v.2.6.3 (Hyatt et al. 2010), and HMMer v.3.3 (http://hmmer.org/). High quality MAGs were annotated using RASTtk v.2.0 (Aziz et al. 2008; Overbeek et al. 2014; Brettin et al. 2015) in order to identify 16S rRNA genes and functional genes of interest. For MAGs that did not have 16S rRNA genes identified by RASTtk, barrnap v.0.9 (https://github.com/tseemann/barrnap) was used to look for short 16S rRNA gene fragments that may have been cut off during contig assembly. Any short 16S rRNA gene fragments identified in the MAG were used to search for the longer accompanying sequence in the raw reads of that sample. Any longer 16S rRNA gene sequence was used in a search query in BLAST (Altschul et al. 1990) to identify the bacterial taxonomy with the highest % identity. Any MAGs that did not recover reliable 16S rRNA were classified using MetaSanity v.1.2.0 (Neely, Graham and Tully 2020). Data availability Sequence data for this project are stored in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) database. 16S rRNA amplicon sequence data for the shipwreck samples can be found under the accession numbers SRR12148233-SRR12148251. Amplicon data for the stainless steel samples and metagenomic data for both can be found under the BioProject accession number PRJNA674937. RESULTS Higher evenness and diversity on steel surfaces compared to sediment Microbial community richness was lower on surface environments (steel, biofoul PVC, and oyster shell) compared to sediment for stainless steel sites with salinity < 10 ppt (Mann-Whitney U test: W = 271.5, P = 0.0153), but there was no difference in richness between surface environments and sediment for sites > 10 ppt (P > 0.05) (Fig. 1). Microbial community evenness was higher on surface environments (stainless steel, ship steel, biofoul PVC, and oyster shell) compared to sediment for stainless steel sites < 10 ppt (Mann-Whitney U test: W = 56, P = 0.0004) and for the shipwreck site (17.9 ppt) (W = 0, P = 0.0003), but there was no difference in evenness at stainless steel sites > 10 ppt (P > 0.05). Overall, Shannon diversity decreased with increasing salinity, regardless of sample type (Linear regression test: df = 70, t = -12.37, P < 0.0001). More specifically, Shannon diversity was higher on steel surfaces compared to sediment and surface controls for the shipwreck site (Mann-Whitney U test: W = 1, P = 0.003), but there was no difference in diversity between surface environments and sediment for stainless steel sites at all salinities. Figure 1. Open in new tabDownload slide Species richness (left), evenness (middle) and Shannon diversity (right) across sample types based on 97% similarity classifications of OTUs from 16S rRNA gene sequencing data. Samples were separated into two groups for analysis: low salinity (< 10 ppt) and high salinity (> 10 ppt) for more accurate fine-scale comparisons between sample types. Microbial community evenness was higher on surface environments (SS, ship, biofoul and oyster shell) compared to sediment for low salinity sites (Mann–Whitney U test: W = 56, P = 0.0004) and for the shipwreck site (W = 0, P = 0.0003). Shannon diversity was higher on steel surfaces than both sediment and surface controls for the shipwreck site only (W = 1, P = 0.003). Figure 1. Open in new tabDownload slide Species richness (left), evenness (middle) and Shannon diversity (right) across sample types based on 97% similarity classifications of OTUs from 16S rRNA gene sequencing data. Samples were separated into two groups for analysis: low salinity (< 10 ppt) and high salinity (> 10 ppt) for more accurate fine-scale comparisons between sample types. Microbial community evenness was higher on surface environments (SS, ship, biofoul and oyster shell) compared to sediment for low salinity sites (Mann–Whitney U test: W = 56, P = 0.0004) and for the shipwreck site (W = 0, P = 0.0003). Shannon diversity was higher on steel surfaces than both sediment and surface controls for the shipwreck site only (W = 1, P = 0.003). Salinity had greatest influence on community composition, followed by substrate Salinity exhibited the strongest correlation with microbial community composition (Adonis test: R2 = 0.275, P = 0.001), followed by substrate (Adonis test: R2 = 0.173, P = 0.001) (Fig. 2). Timescale was also a significant variable, although it presented the weakest correlation (Adonis test: R2 = 0.020, P = 0.006). Microbial communities were not significantly different between 304SS and 316SS, although stainless steel communities as a whole (304SS and 316SS) were significantly different from ship steel communities (ANOSIM test: R = 0.299, P = 0.019). Differences in microbial community composition between stainless steel and ship steel were most strongly influenced at the class level by Gammaproteobacteria (SIMPER analysis: 11.6%), Alphaproteobacteria (11.2%), Betaproteobacteria (10.9%), Planctomycetia (9.0%), unclassified Bacteroidetes (8.9%) and Deltaproteobacteria (7.5%). Figure 2. Open in new tabDownload slide Canonical correspondence analysis (CCA) plot based on 16S rRNA gene sequencing data. Each datapoint represents a unique microbial community composition from one sample. Distance between points represents the degree of difference between community composition of each point. Salinity exhibited the strongest correlation with microbial community composition (Adonis test: R2 = 0.275, P = 0.001). Figure 2. Open in new tabDownload slide Canonical correspondence analysis (CCA) plot based on 16S rRNA gene sequencing data. Each datapoint represents a unique microbial community composition from one sample. Distance between points represents the degree of difference between community composition of each point. Salinity exhibited the strongest correlation with microbial community composition (Adonis test: R2 = 0.275, P = 0.001). Interestingly, differences in community composition between sample types were more evident at higher salinities. Stainless steel and biofoul samples were significantly different from sediment at salinities > 10 ppt (ANOSIM test: R = 0.4872, P = 0.0232), but not at salinities < 10 ppt (Fig. S1, Supporting Information). This suggests that there may be increased microbial selectiveness or specialization due to environment at higher salinities. Additionally, there were two 316SS coupons from two different sites that had visible orange corrosion byproducts, while the rest of the SS coupons exhibited varying degrees of fouling and detritus material but never any orange coloration. These two corroded coupons had no difference in community composition compared to the non-corroded 304SS sample from the same deployment time and site (Fig. S2, Supporting Information). Each pair of 304SS and 316SS samples, regardless of corrosion, grouped together based on site. This suggests that community composition seemed to be the same regardless of the presence or absence of corrosion, while also supporting salinity as the most important factor towards community composition. Clearly, salinity must be taken into account when identifying a core steel microbiome. Core steel microbiome composed of generalist taxa Some microbial taxa were expected to vary in abundance across steel types in accordance with results from a previous study showing higher abundance of corrosion-causing organisms on certain stainless steel types (Garrison, Price and Field 2019). A core microbiome of consistent steel colonizers was expected to emerge as a result of aquatic steel surfaces enabling novel niche establishment. For example, it was expected that core steel microbial taxa may separate based on salinity due to physiological constraints across phylogenies. Results supported this as Betaproteobacteria were more abundant across lower salinity samples (Fig. 3), regardless of sample type, consistent with their known freshwater habitat preference (Wu et al. 2006; Wang et al. 2012; Emerson 2018). There were more Deltaproteobacteria in the sediment samples compared to the steel samples, but only in higher salinity sites. Alphaproteobacteria, Flavobacteria and Sphingobacteriia were more abundant on surface samples (steel and surface controls) compared to sediment samples at higher salinity sites. Low salinity sites showed almost no difference in proportion of dominant taxa across sample types, whereas higher salinity sites showed higher phylogenetic differentiation across sample types. Bacterial genera exhibited similar trends across salinity as those seen at the class level (Fig. S3, Supporting Information). Flavobacterium, Rhodopirellula, Erythrobacter and Hyphomonas were all found at higher proportion on surface environments (steel and surface controls) compared to the sediment at high salinity sites. Figure 3. Open in new tabDownload slide Class-level taxa plot (left) based on 16S rRNA gene sequence classifications. Higher salinity sites (> 10 ppt) showed greater phylogenetic differentiation across sample types than lower salinity sites (< 10 ppt). Alphaproteobacteria, Flavobacteria and Sphingobacteriia were more abundant on the surface samples (steel and surface controls) compared to the sediment samples, but only at higher salinity sites. Stainless steel taxa abundance plotted against ship taxa abundance at the class level (right) indicated that Alphaproteobacteria, Sphingobacteriia, Cytophagia, Anaerolineae and unclassified Proteobacteria were closest to the 1:1 ratio line (with error bars overlapping the line) with > 1% abundance, and thus were most likely to be part of a core steel microbiome. Only stainless steel samples from high salinity sites were used to compare to shipwreck steel samples in this plot. Figure 3. Open in new tabDownload slide Class-level taxa plot (left) based on 16S rRNA gene sequence classifications. Higher salinity sites (> 10 ppt) showed greater phylogenetic differentiation across sample types than lower salinity sites (< 10 ppt). Alphaproteobacteria, Flavobacteria and Sphingobacteriia were more abundant on the surface samples (steel and surface controls) compared to the sediment samples, but only at higher salinity sites. Stainless steel taxa abundance plotted against ship taxa abundance at the class level (right) indicated that Alphaproteobacteria, Sphingobacteriia, Cytophagia, Anaerolineae and unclassified Proteobacteria were closest to the 1:1 ratio line (with error bars overlapping the line) with > 1% abundance, and thus were most likely to be part of a core steel microbiome. Only stainless steel samples from high salinity sites were used to compare to shipwreck steel samples in this plot. A core steel microbiome can be further interpreted by plotting shipwreck taxa abundance against stainless steel taxa abundance to identify taxa with the most similar abundance across steel types (Fig. 3). Datapoints that fell on the 1:1 ratio dashed line, or that have error bars overlapping the line, were determined more likely to be part of a core steel microbiome as opposed to datapoints that did not. Datapoints below the dashed line represent microbes with a higher preference towards stainless steel colonization, while datapoints above the line represent microbes that preferred more traditional steel alloys (i.e. carbon-based or mixed steel alloys less resistant to corrosion). Only high salinity sampling sites (> 10 ppt) were used for core microbiome characterization in order to identify the forces beyond salinity that drive core microbiome composition and function. Alphaproteobacteria, Sphingobacteriia, Cytophagia, Anaerolineae, and unclassified Proteobacteria were the taxonomic classes closest to the 1:1 ratio line with > 1% abundance, and thus most likely to be part of a core steel microbiome (also supported by the taxa plots above showing a higher preference of these taxa on steel surface samples). All of these taxa except Alphaproteobacteria had relatively weak influence on the differences in community composition between stainless steel and ship steel: Sphingobacteriia (SIMPER analysis: 3.7% influence towards community differences), Cytophagia (2.5%), Anaerolineae (2.0%), Alphaproteobacteria (11.2%). At the genus level, unclassified Anaerolineaceae and unclassified Verrucomicrobiaceae were among the closest genera to the core steel dashed line (Fig. S4, Supporting Information). In examining the major outliers from the core steel dashed line, Gammaproteobacteria, Flavobacteria and Betaproteobacteria were more prevalent on the stainless steel coupons, while Deltaproteobacteria, Planctomycetia, and unclassified Bacteroidetes were more prevalent on the shipwreck steel. At the genus level, unclassified Rhodobacteraceae and unclassified Flavobacteriaceae were more prevalent on the stainless steel, while unclassified Planctomycetaceae and unclassified Saprospiraceae were more prevalent on the shipwreck. These results indicate that the core steel microbiome for the steel types and samples used in this study were likely generalist heterotroph microbes that play important supporting roles for specialists that contribute more directly to biocorrosion. A Levins’ niche breadth analysis (Levins 1968) further determined that sediment samples overall exhibited a greater proportion of specialist microbe species compared to surface environment samples (Fig. S5, Supporting Information). A lower niche breadth value indicates a specialist lifestyle which is more selective in its functional role in the environment, whereas a higher niche breadth value indicates a generalist lifestyle which is more likely to use more commonly available sources of energy in a given environment (Levins 1968; Cockburn 1991). All samples had large proportions of high niche breadth values and thus generalist taxa; however, sediment samples simultaneously exhibited significant abundances of low niche breadth values indicative of specialists as well, whereas the steel surface samples did not. This result further supports that generalist microbes likely comprised the majority of the core steel microbiome in this study. An indicator species analysis (De Cáceres and Legendre 2009) supported the niche breadth analysis, also indicating higher proportions of specialist taxa in sediments and high proportions of generalists on steel surfaces. Several genera exhibited both high exclusivity (present only in one sample type and not in other sample types) and high fidelity (present across all samples within one sample type) in the shipwreck sediment samples (indicspecies test: P < 0.05), indicating that these taxa may be unique to this sample type only. These taxa included potential sulfate-reducing genera such as Desulfospira, Desulfatibacillum and Desulfobacterium, as well as the sulfur-oxidizer Thiolapillus. These also included a benzoate-oxidizing sulfate-reducer Desulfocarbo, and an aromatic hydrocarbon-degrader Thalassospira, which could be a result of crude oil remnants from the ship's fuel tank still present in the surrounding sediment. No generalist taxa were identified for indicator species analysis using samples from all salinity values; however, when only including high salinity sites (> 10 ppt), some potential core steel generalists became more evident. Leucothrix was an indicator for 304SS samples (P < 0.05) and is a heterotroph whose filamentous morphology may help form biofilm communities. Sphingobium was an indicator for both 304SS and 316SS samples (P < 0.05), which supports results from the 1:1 ratio plot above (Fig. 3) above as potentially important steel community members. Pelolinea, another filamentous heterotroph, was an indicator for combined ship and ship sediment sample groups (P < 0.05). Five high-quality metagenome assembled genomes (MAGs) were recovered from the steel samples (> 50% complete and < 10% contamination (CheckM v1.0.18 (Parks et al. 2015))) (Table S2, Supporting Information). Two of these MAGs were classified as Rheinheimera sp. and two were classified as Chitinophagaceae. The fifth MAG was classified as Sulfurovum sp., a sulfur-oxidizer. The steel surface associated MAGs exhibited larger average genome size (mean = 2.63 Mbp, standard deviation [SD] = 0.93 Mbp) than surface control MAGs (mean = 1.76 Mbp, SD = 0.58 Mbp), and lower average GC content (mean = 42.3%, SD = 5.0%) than surface control MAGs (mean = 49.2%, SD = 9.4%). These genome characteristics could be inherently associated with core steel microbiomes, although the small number of MAGs recovered limits interpretation, and additional studies may be warranted for further analysis of individual taxa members. Metagenomic analysis also revealed a higher relative abundance of glycoside hydrolase genes, which are indicators for identifying the presence of generalist taxa such as cellulose-, chitin-, and other polysaccharide-degraders (Edwards et al. 2010), on steel samples (mean = 0.22%, SD = 0.03%) compared to surface controls (i.e. oyster shell and biofoul plates) (mean = 0.17%, SD = 0.03%) (Mann–Whitney U test: W = 3, P = 0.066) (Fig. 4). Figure 4. Open in new tabDownload slide Glycoside hydrolase boxplot (left) shows a higher relative abundance of glycoside hydrolase genes on steel samples compared to surface controls (oyster shell and biofoul plates) based on metagenomic sequence data (Mann–Whitney U test: W = 3, P = 0.066). Functional gene bubble plot (right) shows that relative abundance of MIC-related genes such as iron oxidation/reduction and sulfur oxidation/reduction were not significantly affected by salinity (site N11 = 6 ppt, site N3 = 3.6 ppt, shipwreck site = 17.9 ppt). Figure 4. Open in new tabDownload slide Glycoside hydrolase boxplot (left) shows a higher relative abundance of glycoside hydrolase genes on steel samples compared to surface controls (oyster shell and biofoul plates) based on metagenomic sequence data (Mann–Whitney U test: W = 3, P = 0.066). Functional gene bubble plot (right) shows that relative abundance of MIC-related genes such as iron oxidation/reduction and sulfur oxidation/reduction were not significantly affected by salinity (site N11 = 6 ppt, site N3 = 3.6 ppt, shipwreck site = 17.9 ppt). Steel surface functional genes were conserved across salinity while phylogeny diverged Functional genes were hypothesized to follow similar trends to taxonomic data, with varying degrees of influence by salinity, substrate, and timescale. Metagenomic analysis using Hidden Markov model (HMM)-based categorization of biogeochemically-relevant genes (FeGenie (Garber et al. 2020) and LithoGenie (https://github.com/Arkadiy-Garber/LithoGenie)) revealed MIC-related genes (i.e. iron oxidation and reduction, sulfur oxidation and reduction) at similar relative abundances across all stainless steel and ship steel types (Fig. 4, Figs S7 and S8, Supporting Information). Both stainless steel sites used in metagenomic analysis were low salinity sites (6 ppt and 3.6 ppt) compared to the shipwreck site (17.9 ppt), and thus salinity did not have a significant impact on the relative abundance of MIC-related genes. This is in contrast to the 16S rRNA gene sequence results which showed many phylogenetic separations in steel surface community compositions based on salinity. There was also no significant difference in glycoside hydrolase abundance between low salinity stainless steel samples and high salinity ship steel samples, suggesting that functional genes not related to MIC may also be conserved across salinity more so than phylogeny. There was a higher abundance of iron oxidation (Cyc1, Cyc2, FoxABC, FoxEYZ, Sulfocyanin, PioABC, MtoAB), iron reduction (CymA, MtrCAB, OmcF, OmcS, OmcZ, FmnA-dmkA-fmnB-pplA-ndh2-eetAB-dmkB, DFE_0448–0451, DFE_0461–0465, MtrCB, MtrAB, MtoAB-MtrC), sulfur oxidation (DsrABCEFHK, dioxygenase sdo), sulfate reduction (CysN, aprA, sopT, apsK), and oxygen reduction (CytCoxidase coxAB, ccoNOP, ubiquinol oxidase CyoE, oxidase CydAB) proteins found on steel surfaces compared to surface control sample types (although not statistically significant; Mann-Whitney U test: W = 160, P = 0.116), consistent with steel surfaces having a higher functional potential for MIC compared to surface controls (Figs S6–S8, Supporting Information). There were also higher percentages of oxygen reduction related genes in the 304SS, 316SS and biofoul control samples compared to the nearby sediment samples (Fig. S6, Supporting Information) (W = 1, P = 0.19), suggesting an increased use of oxygen as an electron acceptor and therefore indicative of a more oxidative environment rather than a reducing environment on surface samples compared to sediment samples. DISCUSSION The results from this study provide insight into the community composition and functional potential of a core steel microbiome. Bacterial alpha diversity on steel surfaces was inversely related to salinity, as Shannon diversity values increased with decreasing salinity. Furthermore, community evenness was more influential than species richness in determining Shannon diversity. Evenness was not only higher at low salinity sites, but also higher on steel surfaces compared to sediment. This suggests that steel surfaces may exhibit higher bacterial productivity and functional stability than that of nearby sediment (Wittebolle et al. 2009; Haig et al. 2015; Schmidt et al. 2020). This is surprising considering the greater environmental fluctuations seen in the water column compared to the sediment, and this indicates that the stability provided by protective biofilms on steel surfaces must be significant. Evenness can also be used as a signal for differences in core microbiomes (Shade and Handelsman 2012), and thus these results argue that steel surfaces have unique core microbiomes from that of sediment, and that core steel microbiomes will vary based on the environment's salinity. Salinity was the most influential factor for bacterial community composition on steel surface, similar to trends observed in other environment types (Bouvier and del Giorgio 2002; Lozupone and Knight 2007; Herlemann et al. 2011). Interestingly, substrate seemed to become a more significant factor as salinity increased. There was a greater difference in Shannon diversity between substrate types at the shipwreck site, which had the highest salinity (17.9 ppt), compared to stainless steel sites (0.1–15.8 ppt). There was also a greater difference in community composition between substrate types at high salinity stainless steel sites compared to low salinity sites (Fig. 3; Figs S1 and S3, Supporting Information). More specifically, Alphaproteobacteria, Flavobacteria and Sphingobacteriia exhibited greater abundances on surface substrates compared to sediment at higher salinities. This suggests a potential preference of these organisms towards surface attachment and an increased bacterial selectiveness of substrate at higher salinities. The core steel microbiome may have more overlapping community members with sediment and surface control environments as salinity decreases. Heterotrophic generalists consistently emerged as important community members within the core steel microbiome in this study. Higher proportions of generalists were found on steel at high salinities, and several generalist taxa were significant indicators for steel surface substrates. Their taxonomic groups had the most equal percent abundance across different steel types (Fig. 3), and they represented the majority of the MAGs recovered from steel samples. An important biofilm functional gene, glycoside hydrolase, was also found at a higher abundance on all steel surface types compared to surface control groups. These combined results are overwhelming evidence that generalist taxa likely dominate the core steel microbiome. The core steel microbiome taxa at higher salinity sites (>10 ppt) included members of Sphingobacteriia, Cytophagia, Anaerolineaceae, Verrucomicrobiaceae, Chitinophagaceae, and Rheinheimera. The similarity in average abundance of Alphaproteobacteria across steel types supported their role as potential core steel members, however, their high variability and large influence towards community differences (SIMPER analysis: 11.2%) suggest that their core steel members need to be taxonomically refined beyond the class level. Even so, they are among the most phylogenetically diverse classes (Hug et al. 2016), and functional diversity could be beneficial in steel biofilm communities. Sphingobacteriia are known to be important members of aquatic biofilms that degrade biopolymers such as cellulose and chitin that tend to accumulate on these surface environments (Sack, van der Wielen and van der Kooij 2014; Battin et al. 2016). Less is known about Cytophagia functional roles, but they are closely related in phylogeny to Sphingobacteriia. Anaerolineaceae are anaerobic fermenters of a wide variety of carbon sources (McIlroy et al. 2017) and could play an important functional role in the deeper biofilm layers closest to the surface environment. Verrucomicrobiaceae are known to be general heterotrophs and are also capable of degrading biopolymers such as cellulose, chitin and xylan (Cabello-Yeves et al. 2017). Chitinophagaceae and Rheinheimera are both generalist heterotrophs that contribute to, or have genes associated with biofilm production and functional stability (e.g. pilin proteins) (Mai-Prochnow et al. 2004; Sack, van der Wielen and van der Kooij 2014; Schuster and Szewzyk 2016). Recent studies have shown supporting evidence of the importance of generalist taxa on steel biofilm communities. Moura et al. (2018) found increased rates of corrosion with the presence of generalists such as Rhodobacteraceae, Flavobacteriaceae, and Pseudomonadaceae in marine microcosm experiments. Mugge et al. (2019) similarly found increasing abundances of Alphaproteobacteria, Flavobacteriia, Saprospiraceae and Rhodobacteraceae on steel surfaces over time in another marine microcosm corrosion experiment. Steel biofilm communities can often contain carbon-based organic material from many different sources accumulated on the surface environment and thus limit both the attachment ability of other organisms and access to underlying iron sources for growth. Without heterotrophic generalists capable of degrading these potentially complex carbon-based materials, steel biofilms would likely have much lower functional potential. Generalist species have also been shown to be less affected by environmental factors (Pandit, Kolasa and Cottenie 2009), suggesting a potential competitive advantage over specialists in environments that are subjected to constant change such as steel surfaces in estuaries. Steel surfaces found in more static environments, however, could presumably exhibit smaller proportions of generalists and greater proportions of specialists resulting in less biofilm protection and stability, and thus may be an interesting direction for future studies. We also acknowledge that the steel surface community compositions were phylogenetically separated across the spectrum of salinity in this study (0–17.9 ppt), and there were not enough substrate type comparisons to make a strong case for classification of a low salinity core steel microbiome. Additional studies could potentially resolve this with an increased number of steel types sampled from low salinity sites. The higher abundance of glycoside hydrolase genes on steel surface samples compared to surface controls suggests that cellulose-, chitin- and polysaccharide-degraders, which are important members of marine biofilm communities (Christensen et al. 1998; Vu et al. 2009), are also important to steel surface microbial communities and may even have a syntrophic advantage to colonizing alongside corrosion-causing microbes. The oyster shell and biofoul surface controls likely also have biofilms within which these genes would be beneficial, thus the higher abundance of glycoside hydrolase found on steel samples suggests a potential greater need for the gene on steel compared to the surface controls. All of the high-quality MAGs on the steel surface samples, except for the Sulfurovum sp. (sulfur-oxidizer), also contained glycoside hydrolase genes, further supporting these MAGs as being potential core steel microbes, and glycoside hydrolase as an important functional gene for biofilm communities. Other biofilm-related genes were found within the MAGs, such as pilin functional genes, which are critical to surface colonization, biofilm formation, and social interactions with other cells (Tomaras et al. 2003; Mandlik et al. 2008; Anyan et al. 2014). The presence of these genes decreases the possibility that any of these organisms were found on these steel surfaces by chance alone. Potential future studies could clarify further by using a higher number of samples for shotgun metagenomics and/or single cell genomics in order to identify additional genomic trends across steel surface associated MAGs. The steel surface samples also exhibited higher proportions of iron oxidation/reduction and sulfur oxidation/reduction genes compared to surface controls, consistent with these steel surface samples hosting a microbial community that is capable of corrosion metabolisms. The stainless steel samples used for metagenomic analysis were chosen due to the orange rusted appearance of the 316SS coupon, but not on the 304SS coupon, in both cases. Even with the difference in visual appearance, the abundance of iron, sulfur, and glycoside hydrolase functional genes were comparable across 304SS, 316SS and ship steel samples. Both of the visually corroded stainless steel coupons were also from low salinity sites (6 ppt and 3.6 ppt), and as a result, salinity did not seem to have a significant impact on functional potential of steel microbial communities regardless of any biogeochemical processes that were occurring on the steel surface. As seen in the results above, steel surface samples were highly separated based on phylogeny due to both salinity and substrate. However, a large suite of functional genes seemed to be conserved across these salinity and substrate barriers. This is in contrast to previous work showing functional trade-offs and divergence accompanying phylogenetic divergence for microbial communities in the water column across salinity gradients (Dupont et al. 2014). It is possible that steel surface communities exhibit less functional divergence across salinity compared to pelagic communities. This could be due to increased concentrations and more consistent supplies of nutrients on steel surfaces compared to the water column (copiotrophic vs oligotrophic). In addition, the majority of bacterial functional traits have been shown to be largely polyphyletic (Martiny, Treseder and Pusch 2013), and this characteristic may be featured more strongly on these surface communities. The findings in this study support the notion that bacterial functional roles can be both polyphyletic and conserved on steel surface communities. The increased selectiveness inferred across substrates at higher salinities also suggests that freshwater taxa associated with steel surface communities may have a greater number of functional roles compared to higher salinity taxa, or that a greater number of taxa from higher salinities perform the same functional roles within the surface communities. We acknowledge that these results are merely a first step in determining the importance and role of core steel microbiomes, and additional studies focusing on their functional potential across a larger number of samples and within other environment types and steel types will be needed to fully understand submerged steel surface microbiomes and how they might compare to other core microbiomes. The results from this study show that functional potential within these communities can be maintained across salinity, substrate, and timescale. Heterotrophic generalist taxa appear to be important members of the core steel microbiome that likely contribute to the overall formation and stability of these biofilm communities. These submerged steel structures seem to be important participants not only for iron and sulfur cycles, but also for carbon and other heterotrophic nutrient cycling processes in the environment. This study also supports core microbiomes as being important for ecosystem health and function in environments other than the often-referenced host-associated systems. The core steel microbiome ‘behind the scenes’ is not always easily observed compared to microbes associated with visible corrosion by-products, but its role in biofilm formation and functional stability is important for the overall steel community health. As our understanding of these core steel microbiomes improves even further, we can ultimately gain a stronger grasp on MIC prevention and the overall fate of submerged steel structures. ACKNOWLEDGEMENTS We thank Kyra Price and Nathan Richards for their help with sample collection and preliminary studies at the shipwreck site. We also thank the East Carolina University (ECU) Maritime Studies Field School and the Blakeslee and Field laboratories (ECU) for help with sample collection. We thank North Creek Landing Homeowners Association and Matthew's Point Marina for facility access. Sampling permits for the shipwreck were provided by North Carolina Department of Natural and Cultural Resources Permit #17PAS652, NC Office of State Archaeology Permit #17PAS654 and NC Department of Environmental Quality and Coastal Resources Commission Permit #97-17. FUNDING This work was supported by Oak Ridge Associated Universities and East Carolina University Coastal Maritime Council. Conflicts of interest None declared. REFERENCES Aanderud ZT , Jones SE, Fierer N et al. Resuscitation of the rare biosphere contributes to pulses of ecosystem activity . Front Microbiol . 2015 ; 6 : 24 . Google Scholar Crossref Search ADS PubMed WorldCat Alneberg J , Bjarnason BS, De Bruijn I et al. Binning metagenomic contigs by coverage and composition . Nat Methods . 2014 ; 11 : 1144 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat Altschul SF , Gish W, Miller W et al. Basic local alignment search tool . J Mol Biol . 1990 ; 215 : 403 – 10 . Google Scholar Crossref Search ADS PubMed WorldCat Andrews S . FastQC: A Quality Control Tool for High Throughput Sequence Data . 2010 ; Available online at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/. Anyan ME , Amiri A, Harvey CW et al. Type IV pili interactions promote intercellular association and moderate swarming of Pseudomonas aeruginosa . Proc Nat Acad Sci . 2014 ; 111 : 18013 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Aziz RK , Bartels D, Best AA et al. The RAST Server: rapid annotations using subsystems technology . BMC Genomics . 2008 ; 9 : 1 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat Bankevich A , Nurk S, Antipov D et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing . J Comput Biol . 2012 ; 19 : 455 – 77 . Google Scholar Crossref Search ADS PubMed WorldCat Battin TJ , Besemer K, Bengtsson MM et al. The ecology and biogeochemistry of stream biofilms . Nat Rev Microbiol . 2016 ; 14 : 251 . Google Scholar Crossref Search ADS PubMed WorldCat Beech IB , Sunner J. Biocorrosion: towards understanding interactions between biofilms and metals . Curr Opin Biotech . 2004 ; 15 : 181 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat Berg G , Rybakova D, Fischer D et al. Microbiome definition re-visited: old concepts and new challenges . Microbiome . 2020 ; 8 : 1 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat Bouvier TC , del Giorgio PA. Compositional changes in free‐living bacterial communities along a salinity gradient in two temperate estuaries . Limnol Oceanogr . 2002 ; 47 : 453 – 70 . Google Scholar Crossref Search ADS WorldCat Brettin T , Davis JJ, Disz T et al. RASTtk: a modular and extensible implementation of the RAST algorithm for building custom annotation pipelines and annotating batches of genomes . Sci Rep . 2015 ; 5 : 8365 . Google Scholar Crossref Search ADS PubMed WorldCat Cabello-Yeves PJ , Ghai R, Mehrshad M et al. Reconstruction of diverse verrucomicrobial genomes from metagenome datasets of freshwater reservoirs . Front Microbiol . 2017 ; 8 : 2131 . Google Scholar Crossref Search ADS PubMed WorldCat Christensen BB , Sternberg C, Andersen JB et al. Establishment of new genetic traits in a microbial biofilm community . Appl Environ Microbiol . 1998 ; 64 : 2247 – 55 . Google Scholar Crossref Search ADS PubMed WorldCat Cockburn A . An Introduction to Evolutionary Ecology . Oxford : Blackwell , 1991 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Comeau AM , Douglas GM, Langille MG. Microbiome helper: a custom and streamlined workflow for microbiome research . MSystems . 2017 ; 2 : e00127 – 16 . Google Scholar Crossref Search ADS PubMed WorldCat Cáceres MD , Legendre P. Associations between species and groups of sites: indices and statistical inference . Ecology . 2009 ; 90 : 3566 – 74 . Google Scholar Crossref Search ADS PubMed WorldCat Dang H , Lovell CR. Microbial surface colonization and biofilm development in marine environments . Microbiol Mol Biol R . 2016 ; 80 : 91 – 138 . Google Scholar Crossref Search ADS WorldCat Dupont CL , Larsson J, Yooseph S et al. Functional tradeoffs underpin salinity-driven divergence in microbial community composition . PLoS One . 2014 ; 9 : e89549 . Google Scholar Crossref Search ADS PubMed WorldCat Edwards JL , Smith DL, Connolly J et al. Identification of carbohydrate metabolism genes in the metagenome of a marine biofilm community shown to be dominated by Gammaproteobacteria and Bacteroidetes . Genes . 2010 ; 1 : 371 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat Emerson D , Fleming EJ, McBeth JM. Iron-oxidizing bacteria: an environmental and genomic perspective . Annu Rev Microbiol . 2010 ; 64 : 561 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat Emerson D . The role of iron-oxidizing bacteria in biocorrosion: a review . Biofouling . 2018 ; 34 : 989 – 1000 . Google Scholar Crossref Search ADS PubMed WorldCat Feng BW , Li XR, Wang JH et al. Bacterial diversity of water and sediment in the Changjiang estuary and coastal area of the East China Sea . FEMS Microbiol Ecol . 2009 ; 70 : 236 – 48 . Google Scholar Crossref Search ADS WorldCat Garber AI , Nealson KH, Okamoto A et al. FeGenie: A comprehensive tool for the identification of iron genes and iron gene neighborhoods in genome and metagenome assemblies . Front Microbiol . 2020 ; 11 : 37 . Google Scholar Crossref Search ADS PubMed WorldCat Garrison CE , Price KA, Field EK. Environmental evidence for and genomic insight into the preference of iron-oxidizing bacteria for more-corrosion-resistant stainless steel at higher salinities . Appl Environ Microbiol . 2019 ; 85 : e00483 – 19 . Google Scholar Crossref Search ADS PubMed WorldCat Haig SJ , Quince C, Davies RL et al. The relationship between microbial community evenness and function in slow sand filters . mBio . 2015 ; 6 : e00729 – 15 . Google Scholar Crossref Search ADS PubMed WorldCat Helmus MR , Bland TJ, Williams CK et al. Phylogenetic measures of biodiversity . Am Nat . 2007 ; 169 : E68 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat Herlemann DP , Labrenz M, Jürgens K et al. Transitions in bacterial communities along the 2000 km salinity gradient of the Baltic Sea . ISME J . 2011 ; 5 : 1571 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Hol WG , De Boer W, Termorshuizen AJ et al. Reduction of rare soil microbes modifies plant–herbivore interactions . Ecol Lett . 2010 ; 13 : 292 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat Hug LA , Baker BJ, Anantharaman K et al. A new view of the tree of life . Nat Microbiol . 2016 ; 1 : 1 – 6 . Google Scholar Crossref Search ADS WorldCat Hu YO , Karlson B, Charvet S et al. Diversity of Pico-to Mesoplankton along the 2000 km Salinity Gradient of the Baltic Sea . Front Microbiol . 2016 ; 7 : 679 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Hyatt D , Chen GL, LoCascio PF et al. Prodigal: prokaryotic gene recognition and translation initiation site identification . BMC Bioinformatics . 2010 ; 11 : 119 . Google Scholar Crossref Search ADS PubMed WorldCat Jousset A , Bienhold C, Chatzinotas A et al. Where less may be more: how the rare biosphere pulls ecosystems strings . ISME J . 2017 ; 11 : 853 – 62 . Google Scholar Crossref Search ADS PubMed WorldCat Kang DD , Li F, Kirton E et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies . PeerJ . 2019 ; 7 : e7359 . Google Scholar Crossref Search ADS PubMed WorldCat Kip N , Van Veen JA. The dual role of microbes in corrosion . ISME J . 2015 ; 9 : 542 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat Kleindienst S , Grim S, Sogin M et al. Diverse, rare microbial taxa responded to the Deepwater Horizon deep-sea hydrocarbon plume . ISME J . 2016 ; 10 : 400 – 15 . Google Scholar Crossref Search ADS PubMed WorldCat Krueger F . Trim Galore: a wrapper tool around Cutadapt and FastQC to consistently apply quality and adapter trimming to FastQ files . 2012 . Available online at: http://www.bioinformatics.babraham.ac.uk/projects/trim_galore Langmead B , Salzberg SL. Fast gapped-read alignment with Bowtie 2 . Nat Methods . 2012 ; 9 : 357 . Google Scholar Crossref Search ADS PubMed WorldCat Levins R . Evolution in Changing Environments: Some Theoretical Explorations . Princeton, NJ : Princeton University Press , 1968 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Li H , Handsaker B, Wysoker A et al. The sequence alignment/map format and SAMtools . Bioinformatics . 2009 ; 25 : 2078 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Little BJ , Lee JS. Microbiologically Influenced Corrosion. (Vol 3 ) Hoboken, NJ : John Wiley & Sons , 2007 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Lozupone CA , Knight R. Global patterns in bacterial diversity . Proc Nat Acad Sci . 2007 ; 104 : 11436 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat Mai-Prochnow A , Evans F, Dalisay-Saludes D et al. Biofilm development and cell death in the marine bacterium Pseudoalteromonas tunicata . Appl Environ Microbiol . 2004 ; 70 : 3232 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Mandlik A , Swierczynski A, Das A et al. Pili in Gram-positive bacteria: assembly, involvement in colonization and biofilm development . Trends Microbiol . 2008 ; 16 : 33 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat Martiny AC , Treseder K, Pusch G. Phylogenetic conservatism of functional traits in microorganisms . ISME J . 2013 ; 7 : 830 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Matsen FA , Kodner RB, Armbrust EV. pplacer: linear time maximum-likelihood and Bayesian phylogenetic placement of sequences onto a fixed reference tree . BMC Bioinformatics . 2010 ; 11 : 538 . Google Scholar Crossref Search ADS PubMed WorldCat McBeth JM , Emerson D. In situ microbial community succession on mild steel in estuarine and marine environments: exploring the role of iron-oxidizing bacteria . Front Microbiol . 2016 ; 7 : 767 . Google Scholar Crossref Search ADS PubMed WorldCat McBeth JM , Fleming EJ, Emerson D. The transition from freshwater to marine iron‐oxidizing bacterial lineages along a salinity gradient on the Sheepscot River, Maine, USA . Environ Microbiol Rep . 2013 ; 5 : 453 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat McBeth JM , Little BJ, Ray RI et al. Neutrophilic iron-oxidizing “Zetaproteobacteria” and mild steel corrosion in nearshore marine environments . Appl Environ Microbiol . 2011 ; 77 : 1405 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat McIlroy SJ , Kirkegaard RH, Dueholm MS et al. Culture-independent analyses reveal novel anaerolineaceae as abundant primary fermenters in anaerobic digesters treating waste activated sludge . Front Microbiol . 2017 ; 8 : 1134 . Google Scholar Crossref Search ADS PubMed WorldCat Meyer F , Paarmann D, D'Souza M et al. The metagenomics RAST server–a public resource for the automatic phylogenetic and functional analysis of metagenomes . BMC Bioinformatics . 2008 ; 9 : 1 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Moura V , Ribeiro I, Moriggi P et al. The influence of surface microbial diversity and succession on microbiologically influenced corrosion of steel in a simulated marine environment . Arch Microbiol . 2018 ; 200 : 1447 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat Mugge RL , Lee JS, Brown TT et al. Marine biofilm bacterial community response and carbon steel loss following Deepwater Horizon spill contaminant exposure . Biofouling . 2019 ; 35 : 870 – 82 . Google Scholar Crossref Search ADS PubMed WorldCat Nawar F . Biodiversity of Bacteria and Protists Along a Salinity Gradient in the Fraser River Estuary . Doctoral dissertation , Vancouver, Canada : University of British Columbia . 2016 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Neely CJ , Graham ED, Tully BJ. MetaSanity: An integrated microbial genome evaluation and annotation pipeline . Bioinformatics . 2020 ; 36 ; btaa512 . Google Scholar Crossref Search ADS WorldCat Oren A . Diversity of halophilic microorganisms: environments, phylogeny, physiology, and applications . J Ind Microbiol Biot . 2002 ; 28 : 56 – 63 . Google Scholar Crossref Search ADS WorldCat Overbeek R , Olson R, Pusch GD et al. The SEED and the Rapid Annotation of microbial genomes using Subsystems Technology (RAST) . Nucleic Acids Res . 2014 ; 42 : D206 – 14 . Google Scholar Crossref Search ADS PubMed WorldCat Pandit SN , Kolasa J, Cottenie K. Contrasts between habitat generalists and specialists: an empirical extension to the basic metacommunity framework . Ecology . 2009 ; 90 : 2253 – 62 . Google Scholar Crossref Search ADS PubMed WorldCat Parks DH , Imelfort M, Skennerton CT et al. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes . Genome Res . 2015 ; 25 : 1043 – 55 . Google Scholar Crossref Search ADS PubMed WorldCat Pedersen K . Biofilm development on stainless steel and PVC surfaces in drinking water . Water Res . 1990 ; 24 : 239 – 43 . Google Scholar Crossref Search ADS WorldCat Price KA , Garrison CE, Richards N et al. A Shallow water ferrous-hulled shipwreck reveals a distinct microbial community . Front Microbiol . 2020 ; 11 : 1897 . Google Scholar Crossref Search ADS PubMed WorldCat R Core Team . R: A language and environment for statistical computing . 2013 . Sack EL , van der Wielen PW, van der Kooij D. Polysaccharides and proteins added to flowing drinking water at microgram-per-liter levels promote the formation of biofilms predominated by Bacteroidetes and Proteobacteria . Appl Environ Microbiol . 2014 ; 80 : 2360 – 71 . Google Scholar Crossref Search ADS PubMed WorldCat Schloss PD , Westcott SL, Ryabin T et al. Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities . Appl Environ Microbiol . 2009 ; 75 : 7537 – 41 . Google Scholar Crossref Search ADS PubMed WorldCat Schmidt ML , Biddanda BA, Weinke AD et al. Microhabitats are associated with diversity–productivity relationships in freshwater bacterial communities . FEMS Microbiol Ecol . 2020 ; 96 : fiaa029 . Google Scholar Crossref Search ADS PubMed WorldCat Schuster AK , Szewzyk U. Draft genome sequence of Rheinheimera sp. F8, a biofilm-forming strain which produces large amounts of extracellular DNA . Genome Announc . 2016 ; 4 : e00082 – 16 . Google Scholar Crossref Search ADS PubMed WorldCat Shade A , Handelsman J. Beyond the Venn diagram: the hunt for a core microbiome . Environ Microbiol . 2012 ; 14 : 4 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat Shade A , Jones SE, Caporaso JG et al. Conditionally rare taxa disproportionately contribute to temporal changes in microbial diversity . mBio . 2014 ; 5 : e01371 – 14 . Google Scholar Crossref Search ADS PubMed WorldCat Sieber CM , Probst AJ, Sharrar A et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy . Nat Microbiol . 2018 ; 3 : 836 – 43 . Google Scholar Crossref Search ADS PubMed WorldCat Tomaras AP , Dorsey CW, Edelmann RE et al. Attachment to and biofilm formation on abiotic surfaces by Acinetobacter baumannii: involvement of a novel chaperone-usher pili assembly system . Microbiology . 2003 ; 149 : 3473 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat Vandecandelaere I , Nercessian O, Faimali M et al. Bacterial diversity of the cultivable fraction of a marine electroactive biofilm . Bioelectrochemistry . 2010 ; 78 : 62 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat Vu B , Chen M, Crawford RJ et al. Bacterial extracellular polysaccharides involved in biofilm formation . Molecules . 2009 ; 14 : 2535 – 54 . Google Scholar Crossref Search ADS PubMed WorldCat Wang Y , Sheng HF, He Y et al. Comparison of the levels of bacterial diversity in freshwater, intertidal wetland, and marine sediments by using millions of illumina tags . Appl Environ Microbiol . 2012 ; 78 : 8264 – 71 . Google Scholar Crossref Search ADS PubMed WorldCat Willis C , Desai D, LaRoche J. Influence of 16S rRNA variable region on perceived diversity of marine microbial communities of the Northern North Atlantic . FEMS Microbiol Lett . 2019 ; 366 : fnz152 . Google Scholar Crossref Search ADS PubMed WorldCat Wittebolle L , Marzorati M, Clement L et al. Initial community evenness favours functionality under selective stress . Nature . 2009 ; 458 : 623 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat Wu QL , Zwart G, Schauer M et al. Bacterioplankton community composition along a salinity gradient of sixteen high-mountain lakes located on the Tibetan Plateau, China . Appl Environ Microbiol . 2006 ; 72 : 5478 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat Wu YW , Simmons BA, Singer SW. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets . Bioinformatics . 2016 ; 32 : 605 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat Yang B , Wang Y, Qian PY. Sensitivity and correlation of hypervariable regions in 16S rRNA genes in phylogenetic analysis . BMC Bioinformatics . 2016 ; 17 : 1 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2020. Published by Oxford University Press on behalf of FEMS.. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Introducing a “core steel microbiome” and community functional analysis associated with microbially influenced corrosion
JF - FEMS Microbiology Ecology
DO - 10.1093/femsec/fiaa237
DA - 2020-12-30
UR - https://www.deepdyve.com/lp/oxford-university-press/introducing-a-core-steel-microbiome-and-community-functional-analysis-YA09uB3yVj
VL - 97
IS - 1
DP - DeepDyve
ER -