The evolution of bacterial community structure in oil mousses
After the DWH oil spill, studies on bacterial community structures and the active oil degraders have focused mostly on the deepwater oil plume (Hazen et al. 2010; Valentine et al. 2010; Kessler et al. 2011; Redmond and Valentine 2011; Lu et al. 2012; Mason et al. 2012), and oil in sandy beaches or salt marshes (Kostka et al. 2011; Beazley et al. 2012). But an understanding of how bacterial communities evolve in the surface oil during transport from the accident site to salt marshes along the coast remains unclear.
Oceanospirillales, Colwellia, and Cycloclasticus of Gammaproteobacteria dominated the bacterial community in the deepwater oil plume (Hazen et al. 2010; Valentine et al. 2010; Kessler et al. 2011; Redmond and Valentine 2011; Bælum et al. 2012; Lu et al. 2012; Mason et al. 2012). Once oil rose to sea surface (2 km from the accident site), the bacterial community in the oil slick was dominated by 100% of Gammaproteobacteria (93% Pseudoalteromonas), but its percentage decreased to 48% (Pseudomonas, Vibrio, Acinetobacter, and Alteromonas) in oil collected at a station further from the accident site (44 km), along with ~30% of Alphaproteobacteria (Bacteroidetes and SAR 11); temperature fluctuation was attributed as the factor leading to this community shift from deepwater (4°C) to sea surface (20°C) (Redmond and Valentine 2011). Our data showed that Alphaproteobacteria continued to increase to over 60% in the oil mousses collected at stations OSS and CT, 135 and 85 km away from the accident site. Petroleum hydrocarbon analysis suggested that OSS and CT mousses were subjected to moderate weathering, as low-molecular-weight hydro carbons such as n-alkanes (n < 15) and naphthalene homologues dissapeared due to evaporation (Reddy et al. 2011; Liu et al. 2012). Together these data suggest that Alphaproteobacteria became more dominant as the surface oil was weathered gradually during the movement, consistent with laboratory incubation results indicating that Alphaproteobacteria become dominant at the later stage of oil degradation (Röling et al. 2004; Beazley et al. 2012).
Gamma- and Alphaproteobacteria dominated the bacterial communties in both OSS and CT mousses (>90%). The community structures of the two oil mousses were similar in Gammaproteobacteria, but remarkably different in Alphaproteobacteria, which may relate to the hydro carbon composition of the mousse (Fig. 4). The occurrences of Alcanivorax (2–3% OTUs), which are excellent degraders of alkanes including branched ones (Hara et al. 2003; Head et al. 2006), in both OSS and CT oil mousses suggest that these bacteria were degrading alkanes. This argument is supported by the lower ratios of n-C17/pristane and n-C18/phytane and lower concentrations of total alkanes in the CT mousse than the OSS mousse (Fig. 4 and Table 2). Marinobacter and Alteromonas, which are common oil-degrading genera of Gammaproteobacteria (Head et al. 2006), also occurred in the two mousses, representing 15% OTUs. These genera can use a broad range of carbon substrates (Kostka et al. 2011). Consistent with previous work (Redmond and Valentine 2011), Cycloclasticus was not detected in the surface mousses, even though this genus is thought to be the main degrader of PAHs (Harayama et al. 2004; Head et al. 2006; Hazen et al. 2010; Bælum et al. 2012), and can grow in both 4°C and 20°C (Coulon et al. 2007). This phenomenon may be explained by high sea surface temperatures at stations OSS (28.7°C) and CT (28.4°C) when samples were collected, which may be too high for Cycloclasticus growth. More research is needed to test the optimum temperature range for Cycloclasticus.
Table 2. Concentrations of petrolume hydrocarbons in oil mousses collected at stations OSS, CT, and MP, and surface sediments at stations SG and SC (data adapted from Liu et al. 2012)
Belonging to Alphaproteobacteria, Erythrobacter appeared with 6% OTUs in OSS mousse, but only 1% in CT mousse. Erythrobacter, anoxygenic phototrophic bacteria (Kolber et al. 2001), may contribute to aromatic hydrocarbon degradation in oil-contaminated beaches (MacNaughton et al. 1999; Chung and King 2001; Röling et al. 2004; Beazley et al. 2012). Erythrobacter grew well in OSS mousse due to strong irradiance on sea surface in the Gulf and ample oil substrates, but its percentage decreased in CT mousse, perhaps caused by further decrease of n-alkane and aromatic hydrocarbon contents in the mousse (Fig. 4). Another dominant Alphaproteobacteria genus in the OSS mousse was Rhodovulum, accounting for 17% of the total OTUs. In tropical waters, two Rhodovulum-related strains, which can degrade petroleum aromatics, were identified (Teramoto et al. 2010). As mentioned above, the high surface water temperatures in the northern Gulf of Mexico were optimal for the growth of Erythrobacter and Rhodovulum (Koblížek et al. 2003; Kumar et al. 2008). Stappia of Rhodobacterales, representing 6% of total OTUs in both OSS and CT mousses, can degrade both aliphatic and aromatic petroleum hydrocarbons under relatively high temperature (Al-Awadhi et al. 2007; Coulon et al. 2007). High abundance of Thalassospira bacteria occurred in the oil mousse, especially in CT oil mousse with 23% of total OTUs. The mesophilic Thalassospira can use aromatics, such as naphthalene, dibenzothiophene, phenanthrene, and fluorene, efficiently as substrates (Kodama et al. 2008). The Rhizobiales order was also abundant in OSS and CT mousses (10–15% of total OTUs), consisting of genera Bartonella, Methylobacterium, Agrobacterium, and Parvibaculum, the degraders of aromatics (Wang et al. 2008). We conclude that the dominance of these Alphaproteobacteria in the two mousses may relate to strong irradiance and high temperature in Gulf surface waters.
Alphaproteobacteria might be important in degrading the aromatic hydrocarbons, since Erythrobacter, Rhodovulum, Stappia, and Thalassospira can degrade aromatics, as described above. Consistently, concentrations of PAHs and alkylated PAHs were four to eight times lower in CT oil mousse than in OSS mousse (Fig. 4), as CT mousse was more weathered than the OSS one (Liu et al. 2012). Moreover, the ratios of phenanthrene/chrysene, an indicator of oil degradation (Pastor et al. 2001), were 1.0, 0.3, and 0.1 in OSS, CT, and MP mousse, respectively (Table 2), supporting the important role of these Alphaproteobacteria in degrading aromatic hydrocarbons after the first wave of degradation, when small-chained n-alkanes were lost already due to evaporation (Liu et al. 2012). More research on metabolic genes is needed to decipher the direct role of these bacteria in the degradation process.
The bacterial community in MP mousse consisted of some common oil degraders found in OSS and CT mousses, including Thalassospira, Stappia, Erythrobacter, Alcanivorax, and Marinobacter, as expected. However, two unique genera occurred in the MP mousse. Arcobacter, belonging to Epsilonproteobacteria, was found in 7.8% of total OTUs, and this genus may relate to sulfide oxidation in oil-contaminated environments under microaerophilic conditions (Voordouw et al. 1996; Gevertz et al. 2000; Wirsen et al. 2002). The relatively high abundance of Arcobacter in MP mousse suggests that this marsh system is highly eutrophic, and these bacteria might originate from the suboxic or anoxic sediment and migrate to the oil by sediment resuspension. The Vibrio genus was remarkably abundant, representing 57% of the total OTUs in MP mousse. Consistently, the abundance of Vibrio vulnificus was 10–100 times higher in tars balls washed onto marshes and beaches along the coast of Mississippi and Alabama after the DWH oil spill than that in seawater and sands (Tao et al. 2011). Vibrio was identified in surface mousse both near the accident site, sandy beaches, and marsh sediments, but only as a minor component (Kostka et al. 2011; Redmond and Valentine 2011; Beazley et al. 2012). Vibrio is often a dominant species in eutrophic salt marsh systems (Ansede et al. 2001). Therefore, the dominance of Vibrio in MP mousse but not in the other samples suggests that these indigenous bacteria of salt marshes may degrade aromatic and perhaps other hydrocarbons actively (West et al. 1984; Hedlund and Staley 2001; Thompson et al. 2004; Beazley et al. 2012). This argument is supported by the three-time decrease of total PAH levels from CT to MP mousse (Fig. 4). Vibrio could have been simply accumulated in the mousse, but it is more plausible that Vibrio became involved in oil degradation to take advantage of the carbon substrate, considering its dominance in the community structure. The bacterial community, including Proteobacteria, Actinobacteria, Bacteroidetes, and Firmicutes, was more diversified in sediments of coastal salt marshes impacted by the DWH oil spill (Beazley et al. 2012), which resembles those in sediment contaminated by the oil (later) more than those in the MP oil mousse collected from the marsh grass. This finding suggests the influence of indigenous bacterial communities on development of the oil degraders.
Bacterial community structures in sediments and overlying water
Results from the sediments and overlying waters sampled 1 year after the DWH oil spill suggest that the observed bacterial communities were involved with later stages of biodegradation. Bacterial community diversity tends to decrease at the initial stages of biodegradation, followed by an increase at later stages in laboratory petroleum incubations (Röling et al. 2004). The high Shannon indices in SG and SC sediments (Table 1) suggest that the bacterial communities may have recovered partially after the initial oil pulse, as the dominant species Oceanospirillales, Colwellia, and Cycloclasticus in the deepwater plume (Hazen et al. 2010; Redmond and Valentine 2011), constituted only minor percentages in the sediment. It is also possible that bacterial communities in sediments were different and more complex than those in the water column at the beginning of the biodegradation. Alpha- and Gammaproteobacteria together accounted for 40–50% of the total OTUs, but no species was as dominant as those in the oil mousses. Pseudomonas of Gammaproteobacteria represented 5% and 14% of the SG and SC communities, respectively. In addition, Hyphomicrobium and Rhodovibrio of Alphaproteobacteria each accounted for about 2% of the sediment bacterial communities, and these species are known to relate to oil contamination in marine environments (Head et al. 2006; Wu et al. 2009; Kostka et al. 2011). Presumably they were degrading oil actively, since considerable oil, including relatively labile low-molecular-weight n-alkanes, aromatics and BTEX (benzene, toluene, ethylbenzene, and p-, m-, and o-xylenes), remained in these sediments one year after the spill (Liu et al. 2012).
Methylococcus and Methylobacter of Gammaproteobacteria accounted for 3–7% of the bacterial communities in SG and SC sediments and SC overlying waters. The occurrence of these type I methanotrophs suggests the existence of low concentrations of methane (Hanson and Hanson 1996). A fraction of methane might be trapped in the oil that was deposited in the deep-sea sediment under the low temperature and high pressure after the DWH oil spill, considering that over 50% of the crude Macondo oil was methane (Ryerson et al. 2011). In addition, methane may have been produced from anaerobic degradation of petroleum hydrocarbons (Widdel and Rabus 2001). This argument is supported by the presence of 1% Desulfobacterium sp., a sulfate-reducing bacterium (Egli et al. 1987), in the bacterial communities of both SG and SC sediments.
A smaller Shannon Index (4.4) suggests that the bacterial community in the overlying water was not as diversified as in the sediments (Table 1). The oil on surface sediment may have affected the bacterial community in the overlying water. For example, the oil degraders Vibrio harveyi and Brevundimonas sp. accounted for 4% and 11% of the community in the overlying water, respectively (Harwati et al. 2007; Kryachko et al. 2012). Ralstonia of Betaproteobacteria, which can degrade hydrocarbons (Płaza et al. 2008), dominated the bacterial community (47% of total OTUs). The Ralstonia strain can degrade BTEX as the sole carbon source (Lee and Lee 2001). Considering that appreciable quantities of BTEX remained in the sediment (1–2 μg/g) (Liu et al. 2012), perhaps these relatively soluble aromatic may diffuse to the overlying water slowly and be degraded by Ralstonia. The excitation-emission matrix (EEM) analysis on the overlying water showed a clear oil-related contour (Ex 270–290 nm, Em 330–360 nm, data not shown), which resembles an oil component (C2) of the crude oil (Zhou et al. 2013). Interestingly, Saccharophagus degradans, a cellulose degrader (Zhang and Hutcheson 2011), represented 9% of the bacterial community in the overlying water. One possible explanation for the growth of this bacterium is that oil degraders in sediment produced polysaccharide-related biosurfactants (Desai and Banat 1997), such as rhamnolipids by Pseudomonas. Perhaps biosurfactants were degraded by these bacteria after they diffused into the overlying water. More research is warranted for testing this speculation.
Oil seeps are widespread in the Gulf of Mexico, so it is not surprising that indigenous bacteria have a strong potential to degrade hydrocarbons (MacDonald et al. 1989). In methane and sulfate-rich cold seeps, the microbial communities in surface sediments (~6 cm) were dominated by Epsilon- and Deltaproteobacteria (Mills et al. 2003; Reed et al. 2006). Beggiatoa mat (Gammaproteobacteria) can also dominate surface sediments of cold hydrocarbon seeps (Sassen et al. 1993). These bacterial communities differ from those in SG and SC sediments. However, bacterial communities in certain cold oil seeps resemble those in the SG and SC sediments. For example, relatively high abundance of Pseudomonas of Gammaproteobacteria, Actinobacteria, and low G+C Firmicutes (33–36% each) were identified at a station with extensive gas hydrates in the northern Gulf of Mexico (Lanoil et al. 2001). Similarly, relatively high percentages of Actinobacteria and Firmicutes (4–8% each) occurred in SG and SC sediments. Chlorofexi/green non-sulfur bacteria were identified in high abundance (12% clones) in the bacterial communities from the cold-seep sediments from Florida Escarpment (Reed et al. 2006). Likewise, we identified about 4% OTUs as Chlorofexi, suggesting that bacterial activities in the SG and SC sediments resemble certain oil-seep sediments (Morris et al. 2004).