Distinct Bacterial Communities in Surficial Seafloor Sediments Following the 2010 Deepwater Horizon Blowout

Abstract

A major fraction of the petroleum hydrocarbons discharged during the 2010 Macondo oil spill became associated with and sank to the seafloor as marine snow flocs. This sedimentation pulse induced the development of distinct bacterial communities. Between May 2010 and July 2011, full-length 16S rRNA gene clone libraries demonstrated bacterial community succession in oil-polluted sediment samples near the wellhead area. Libraries from early May 2010, before the sedimentation event, served as the baseline control. Freshly deposited oil-derived marine snow was collected on the surface of sediment cores in September 2010, and was characterized by abundantly detected members of the marine Roseobacter cluster within the Alphaproteobacteria. Samples collected in mid-October 2010 closest to the wellhead contained members of the sulfate-reducing, anaerobic bacterial families Desulfobacteraceae and Desulfobulbaceae within the Deltaproteobacteria, suggesting that the oil-derived sedimentation pulse triggered bacterial oxygen consumption and created patchy anaerobic microniches that favored sulfate-reducing bacteria. Phylotypes of the polycyclic aromatic hydrocarbon-degrading genus Cycloclasticus, previously found both in surface oil slicks and the deep hydrocarbon plume, were also found in oil-derived marine snow flocs sedimenting on the seafloor in September 2010, and in surficial sediments collected in October and November 2010, but not in any of the control samples. Due to the relative recalcitrance and stability of polycyclic aromatic compounds, Cycloclasticus represents the most persistent microbial marker of seafloor hydrocarbon deposition that we could identify in this dataset. The bacterial imprint of the DWH oil spill had diminished in late November 2010, when the bacterial communities in oil-impacted sediment samples collected near the Macondo wellhead began to resemble their pre-spill counterparts and spatial controls. Samples collected in summer of 2011 did not show a consistent bacterial community signature, suggesting that the bacterial community was no longer shaped by the DWH fallout of oil-derived marine snow, but instead by location-specific and seasonal factors.

Keywords: Cycloclasticus; Deepwater Horizon; MOSSFA; bacterial populations; marine sediment; marine snow.

FIGURE 1

Oil-contaminated sediment cores, characterized by a red-brown layer at the top of the gray sediment. The photos show a representative core from the RV Oceanus cruise in September 2010 (A), core GIP16 collected in Mid-October 2010 (B), core MUC19 collected at the end of November 2010 (C), and core E01801 collected in July 2011 (D). Close-up of oil-derived marine snow flocs collected from the surface of 2010 September sediment, and spread in a Petri dish, with small dark oil droplets in millimeter size (E). Surface of MUC 19 without and with UV illumination, showing the bare surface of the sediments, and the green fluorescence sheen under UV, indicating petroleum hydrocarbons (F). Seafloor map of sampling sites around the Macondo wellhead (G).

FIGURE 2

Phylum-level resolution pie chart plots of bacterial 16S rRNA gene clone libraries and reverse-transcript 16S rRNA clone libraries from oil polluted sediment samples (labeled with ^∗) and control sites.

FIGURE 3

Order, family- and genus-level identification of 16S rRNA clones representing the Alphaproteobacteria and Deltaproteobacteria in the sequence dataset shown in Figure 2 at phylum-level resolution. The y-axis shows the total number of sequences. The color key identifies the 16S rRNA gene contributions for every sediment sample. Yellow, orange, red and brown colors were chosen for oil-contaminated samples, and green and blue colors were chosen for control samples lacking visible oil impact.

FIGURE 4

16S rRNA gene phylogenetic tree of the deltaproteobacterial families Desulfobacteraceae and Desulfobulbaceae. The scale bar corresponds to 10% sequence distance (Jukes-Cantor). The phylogeny was rooted with the Gammaproteobacterium Colwellia psychrerythraea.

FIGURE 5

Order, family- and genus-level identification of 16S rRNA clones representing the Gammaproteobacteria in the sequence dataset shown in Figure 2 at phylum-level resolution.

FIGURE 6

16S rRNA gene phylogenetic tree of Cycloclasticus sequences derived from both water column and the surficial sediment of the Gulf of Mexico at various time points. The orange-colored clone is from May 2010 surface oil slick; red clones are from post-plume water (Valentine et al., 2010; Redmond and Valentine, 2012; Yang et al., 2016); blue clones are from surficial sediments of September, October and November 2010; green clones represent phylotypes and a new strain from oil slick incubations (Gutierrez et al., 2013) and 4°C enrichments (Yang, 2014).

FIGURE 7

Order, family- and genus-level identification of 16S rRNA clones representing the phyla Bacteroidetes, Planctomycetes, and Verrucomicrobia in the sequence dataset shown in Figure 2.

FIGURE 8

Principal coordinates analysis (PCoA) of 16S rRNA gene distances between phylotypes in all sediment samples. Solid symbols stand for the DWH-impacted sediments; open symbols represent the control sediments without visible imprint of oil-derived sedimentation, or samples below the oil-impacted sediment surface, as in the case of GIP16 3–4 cm.