SPECIAL ISSUE ON THE GULF OF MEXICO RESEARCH INITIATIVE: TEN YEARS OF OIL SPILL AND ECOSYSTEM SCIENCE

GoMRI Insights into Microbial Genomics and Hydrocarbon Bioremediation Response in Marine Ecosystems

By Shannon Weiman, Samantha B. Joye, Joel E. Kostka, Kenneth M. Halanych, and Rita R. Colwell

Light microscope image (10×) of oil droplets (large round shapes) with bacteria visible inside and outside of the oil. A pure culture of “Candidatus Macondimonas diazotrophica,” a novel, nitrogen-​fixing crude oil degrader taxon, was recovered from Florida beach sands contaminated with Macondo oil from the Deepwater Horizon discharge. Image credit: Smruthi Karthikeyan, Georgia Institute of Technology

124 Oceanography | Vol.34, No.1 ABSTRACT. The Deepwater Horizon oil spill represents one of the most damag- tions that is widely translatable to other ing environmental catastrophes of our generation. It contaminated vast areas of the systems. Lessons learned from GoMRI open ocean, the deep sea, and the shoreline of the Gulf region and disrupted its ecosys- research form a foundation for under- tems, with both residual and long-term impacts. At the core of all of these ecosystems standing how microbes in various ecosys- are microbial communities that perform essential biogeochemical processes and eco- tems around the globe respond to diverse system services such as carbon and nutrient cycling. Despite their importance, rela- environmental disturbances. From the tively little was known about marine microbes that degrade hydrocarbons in the Gulf Arctic to the equator and the deep sea to of Mexico prior to the Deepwater Horizon spill, nor the effect of hydrocarbons on the coastal shores, these principles apply to microbiology of the Gulf region. Research carried out through the Gulf of Mexico describing how microbes maintain and Research Initiative (GoMRI) revealed cooperative microbial communities operating restore ecosystem balance. These insights at the heart of bioremediation services with highly adaptive and complex dynamics. In will help scientists better understand and addition, these efforts established new methods for assessing and monitoring ecosys- prepare for future catastrophes, from a tem health, whereby microbial population genetics can serve as indicators of biogeo- tanker spill to the long-term disruptions chemical disruptions and/or restoration status in marine and coastal environments. of climate change. Although much research is still needed to fully understand and engage microbially In summary, the research carried out mediated bioremediation services, GoMRI constructed a strong foundation of meth- by GoMRI investigators led to new meth- ods, discoveries, and overarching principles to build upon. These insights and tools will ods for monitoring and assessing ecosys- help scientists better prepare for, and respond to, future environmental catastrophes, tem health using microbial populations from oil tanker spills to long-term disruptions of climate change. as ecosystem indicators in marine and coastal environments. Scientists can now INTRODUCTION provide resources for advancing under- deploy omics tools to take the pulses of One of the most damaging environmen- standing of marine hydrocarbon micro- microbial communities, identify distur- tal catastrophes of our generation, the biology among other Gulf biological, bances, and guide mitigation strategies— Deepwater Horizon (DWH) oil spill dis- physical, and chemical components. The notably at early stages before widespread charged 4.9 million barrels of oil and microbiology research included devel- damage occurs. Although much work 250,000 metric tonnes of natural gas into opment and application of genomic and is still needed to fully understand and the Gulf of Mexico and contaminated vast bioinformatics tools that enabled scien- engage microbially mediated biogeo- areas of the open ocean, the deep sea, and tists to examine hydrocarbon-degrading​ chemical processes that underpin ocean the shoreline. Its disruption of ecosystems microbes in the context of complex micro- systems and their resilience, GoMRI carries residual and long-lasting impacts. bial communities and in unprecedented research has prepared future generations At the core of all ecosystems are detail. With an array of transformational to better protect and restore invaluable microbial communities that constitute “omics” tools, scientists gained valuable habitats by guiding disaster preparedness the foundation for all life by perform- insight into how microbes respond to and response in the face of diverse envi- ing essential services such as carbon and hydrocarbon infusions and restore eco- ronmental stressors and natural disasters nutrient cycling, as well as other core system health. This research discovered around the globe. biogeochemical processes. Microbes are novel species, genes, metabolic pathways, Earth’s first responders, and by rapidly and community dynamics that are instru- OMICS TECHNOLOGIES reacting and adapting to changing con- mental to hydrocarbon degradation and ENABLE NEW BIOGEOCHEMICAL ditions, they restore balance and stabil- related ecosystem functions. In combina- DISCOVERIES ity to the entire ecosystem. In the context tion, the findings revealed an extensive Discovery of New Oil-​Degrading of oil spills, microbes serve as emergency and intricate natural capacity of microbes Microbial Species, Genes, cleanup crews by feeding on hydro- in the Gulf of Mexico to catalyze bioreme- and Enzymes carbons and contributing other funda- diation of petroleum hydrocarbons. This A powerful toolbox of omics technologies mental remediation services. new information will guide future mitiga- and bioinformatic methods, including Despite their importance, relatively tion and restoration strategies to harness genomics, transcriptomics, proteomics, little was known about hydrocarbon- the natural bioremediation capabilities of metabolomics, and metagenomics, is now degrading marine microbes resident in microorganisms without further disturb- available that allows scientists to probe the Gulf of Mexico prior to the DWH ing sensitive ecosystems. the structures and functions of microbial spill or about the effects of hydrocar- Broadly, the GoMRI researchers have communities that form the foundation of bons on the microbiology of the Gulf provided an outline of core ecological marine ecosystems (Grossart et al., 2020). region. The Gulf of Mexico Research and evolutionary principles regarding These technologies provide research- Initiative (GoMRI) was established to microbial response and ecosystem func- ers with the tools to comprehensively

Oceanography | March 2021 125 examine microbial dynamics in real- sense and respond to the oil, using innate metabolites among species, collabora- world situations (Mason et al., 2014a,b; mobility to pursue oil via flagella-based tive networks were created that together Rodriguez-R et al., 2015). The advanced motility and chemotaxis. They are stim- completed the oil degradation process. techniques and bioinformatics methods ulated to express hydrocarbon degrada- In other cases, researchers determined used generated discoveries and impacts tion genes that enable metabolism of var- that certain species specialize in breaking far beyond what was previously thought ious hydrocarbon compounds as energy down the byproducts that are generated possible (Karthikeyan et al., 2019). sources. These hydrocarbon utilizers by other species. Together, these microbial GoMRI researchers discovered many reproduce rapidly in the presence of oil to networks can accomplish far more, with new microbial taxa that thrive in oil-​ dominate the microbial ecosystem in con- greater efficiency, than any single species contaminated marine environments, taminated waters and sediments, creat- could. Taking such a holistic approach, along with novel genes and metabolic ing a microbial bloom (Kostka et al., 2011; the research done by the GoMRI scien- pathways they use for hydrocarbon deg- Doyle et al., 2018, 2020; Karthikeyan et al., tists has illuminated collective metabolic radation (see image on opening page of 2019). Many oil-degrading microbes pos- pathways and ecosystem functions that this article; Hazen et al., 2010; Kostka sess multiple pathways for hydrocarbon would have been overlooked by examin- et al., 2011; Mason et al., 2014a; Crespo- degradation, and which metabolic path- ing only individual species (Mason et al., Medina et al., 2014; Rodriguez-R et al., way is induced may depend on environ- 2012; Doyle et al., 2020). 2015; Kleindienst et al., 2015a; Yang et al., mental conditions and type of exposure In short order, the DWH spill caused 2016; Karthikeyan et al., 2019). In addi- (Karthikeyan et al., 2020a). major shifts in microbial ecosystem tion, major known classes of microbes, Microbes often exhibit cooperative structure and function. Rare biosphere for example, Bacteroidetes, were found to metabolism by partitioning metabolic species capable of degrading hydrocar- harbor a previously unrecognized poten- pathways among community members bons rapidly outgrew neighbors, ulti- tial for hydrocarbon biodegradation (Liu (Zengler and Zaramela, 2018). These mately accounting for up to 90% of the and Liu, 2013). However, the many major microbial partnerships or collaborations community (Kleindienst et al., 2015a; players and ecotypes discovered repre- are critical to microbial community func- Karthikeyan et al., 2019, 2020a). At the sent only a fraction of all the microbes tions and processes, including hydro- same time, microbes that typically dom- involved in hydrocarbon degradation carbon degradation (Joye et al., 2016). inate healthy environmental systems, (Gutierrez et al., 2018). Microbes can also create new catabolic such as ammonia oxidizers and other pathways by shuffling hydrocarbon deg- autotrophs and heterotrophs, declined Oil-Degrading Microbial radation genes between and among com- in relative abundance. Coastal nitrifier Collaboration and Community munity members via lateral gene trans- populations in offshore pelagic waters Dynamics fer, generating novel combinations of underwent alteration in species com- Oil-degrading microbes are ubiquitous enzymes and metabolic capabilities position for at least a year, returning to in marine and estuarine environments in new hosts. normal only after the oil had dissipated. around the world, but their abundance is GoMRI research led to the discovery (Newell et al., 2014; Huettel et al., 2018) low when oil is absent. As part of the “rare of metabolic partnerships for oil degra- Community composition shifted over biosphere,” they harbor metabolic poten- dation. Metagenomic analysis revealed time as hydrocarbon oxidizing bacterial tial of ecological significance; they act as that genes within a single metabolic path- blooms consumed available hydrocar- first responders in the event of an oil spill, way may be distributed throughout the bons, and then declined, being replaced providing critical ecosystem cleanup community, rather than contained within by microbial species capable of degrad- and stabilization functions that impact a single microbial species or strain. In ing the residual metabolic byproducts the fate and environmental distribution some cases, different bacterial species (Figure 1; Kostka et al., 2011; Rodriguez-R of hydrocarbons. were found to specialize in specific met- et al., 2015; Kamalanathan et al., 2020). When oil spills occur, microorganisms abolic steps, but because they shuttled Such microbial succession patterns have

FIGURE 1 (NEXT PAGE). Microbial community functional shifts in coastal sediments in response to oil from the Deepwater Horizon discharge. Genes coding for selected molecular functions related to hydrocarbon degradation, nutrient scavenging and response, photosynthesis, and housekeeping genes are listed (left), with mean genome equivalents per group of samples (middle), and log2 of pre-oil/oiled and oiled/recovered (recov.) fold changes (right). Gene abundance was assessed as the average genome equivalents (mean copies per bacterial cell) at each sampling period. The log2 fold change was estimated as the log2 of the ratio of normalized counts between pre-oiled samples (S1 to S4) and oiled samples (A to G) and between oiled samples and recovered samples (I600, I606, J598, and J604). P values were estimated using a negative binomial test. Interesting patterns were the succession of genes related to easily degradable, light-hydrocarbon fractions, followed by genes specializing in polycyclic aromatic hydrocarbons (PAHs) and complex aromatics, and an increase in nitrogen fixation genes that denote nitrogen limitation for the microbial communities during crude oil biodegradation. CoA = coenzyme A. Figure from Rodriguez-R et al. (2015)

126 Oceanography | Vol.34, No.1 Abundance More abundant in Gene ontology term GO ID across time points oil preoil recov. oil alkane 1-monooxygenase 0018685 rubredoxin-NAD+ reductase 0015044 Alkane rubredoxin-NADP reductase 0015046 ferredoxin-NAD+ reductase 0008860 0008726 Initial oxidation alkanesulfonate monooxygenase 0004022 alcohol dehydrogenase (NAD) Alcohol oxidation 0004029 Aldehyde oxidation aldehyde dehydrogenase (NAD) 0033721 alkanal monooxygenase (FMN-linked) Fatty acid-CoA 0047646 fatty-acyl-CoA synthase 0004321

Degradation of a liphatics Degradation acyl-CoA oxidase 0031955 Fatty acyl-CoA acyl-CoA dehydrogenase 0003997 medium-chain-acyl-CoA dehydrogenase 0003995 long-chain-acyl-CoA dehydrogenase 0070991 Enoyl-CoA very-long-chain-acyl-CoA dehydrogenase 0004466 isovaleryl-CoA dehydrogenase 0017099 enoyl-CoA hydratase 0008470 0004300 long-chain-enoyl-CoA hydratase 0016508 Hydroxy-acyl-CoA palmitoyl-CoA oxidase 0016401 3-hydroxyacyl-CoA dehydratase 0018812 3R-hydroxyacyl-CoA dehydratase

Beta-oxidation 0080023 Oxo-acyl-CoA 3-hydroxyacyl-CoA dehydrogenase 0003857 acetyl-CoA C-acyltransferase 0003988 Acetyl-CoA phenol 2-monooxygenase 0018662 2,4-dichlorophenol 6-monooxygenase 0018666 Phenols biphenyl-2,3-diol 1,2-dioxygenase 0018583 catechol 2,3-dioxygenase 0018577 phenylacetone monooxygenase 0018695 4-hydroxyacetophenone monooxygenase 0018456 Ketones 0033776 cyclohexanone monooxygenase 0033767 benzene 1,2-dioxygenase 0018667 Benzenes toluene dioxygenase 0018619 biphenyl 2,3-dioxygenase 0018624 benzoate 1,2-dioxygenase 0018687 Benzoates 2-chlorobenzoate 1,2-dioxygenase 0018623 4-hydroxybenzoate octaprenyltransferase 0018626 0018618 terephthalate 1,2-dioxygenase 0008412 Other 3-carboxyethylcatechol 2,3-dioxygenase 0018628 2,6-dioxo-6-phenylhexa-3-enoate hydrolase 0047070 carboxylic dihydroxy-dihydro-p-cumate dehydrogenase 0018774 3-oxoadipate CoA-transferase 0018511 acids 0047569

aromatics of aromatics Degradation diphenols oxidoreductase (acceptor: oxygen) o-succinylbenzoate-CoA ligase 0016682 naphthalene 1,2-dioxygenase 0008756 0018625 cyclopentanol dehydrogenase 0055041 cis-2,3-dihydrobiphenyl-2,3-diol dehydrogenase 0018509 coniferyl-aldehyde dehydrogenase 0050269 4-hydroxyphenylacetate 3-monooxygenase 0052881 Aromatics cyclopentanone monooxygenase 0047799

Nutrient nitrate transmembrane transporter 0015112 N urea transmembrane transporter 0015204 scavenging nitrogenase 0016163 & response allantoicase 0004037 chromate transmembrane transporter 0015109 siderophore transmembrane transporter 0015343 Fe ferrous iron binding 0008198 iron-responsive element binding 0030350

e– transporter, cyclic e– transport pathway 0045156 Photosynthesis ribulose-bisphosphate carboxylase 0016984 chlorophyll binding 0016168

phenylalanine-tRNA ligase 0004826 Housekeeping structural constituent of ribosome 0003735 genes phosphoglycerate kinase 0004618 3'-5' exonuclease 0008408

Preoiled / May 2010 (S1, S2, S3, S4) 10-2 10-1 1 10 100 -10 -5 0 5 -5 0 5 10 Oiled / Jul 2010 (A, B, C) ● p-value adjusted ≤ 0.01 Oiled / Oct 2010 (E, F, G) p-value adjusted ≤ 0.05 Genome equivalents ● Log2 fold-change Recovered / Jun 2011 (I600, I606, J598, J604) ● p-value adjusted > 0.05 (mean copies per cell)

Oceanography | March 2021 127 been documented across many different In the deep-sea plume, methane, and are known generally to play import- ecosystems, each with unique species and potentially propane, butane, and ethane, ant roles in anaerobic methane oxida- dynamics, driven by abundance of dif- fueled the early microbial response tion and anaerobic hydrocarbon (oil) ferent hydrocarbon fractions and related (Valentine et al., 2010; Joye et al., degradation, they were not major play- environmental factors and at different 2011; Crespo-Medina et al., 2014). ers in pelagic or benthic oil or gas deg- time points. Thus, conclusions can be Methane was a major driver in the radation during DWH. Immediately drawn from integrating data across time DWH spill, spawning the emergence of after the spill, when n-alkanes and and localities but must be contextual- Methylomonas species with methane oxi- cycloalkanes were more abundant, ized when making broader observations dation capabilities (Figure 2; Crespo- Oceanospirillaceae and Pseudomonas spp. (Crespo-Medina et al., 2014). Medina et al., 2014). Although Archaea were dominant (Figure 3; Hazen et al., 2010). These species were later sup- planted mainly by Colwellia spp. and, to a lesser degree, Cycloclasticus and 0 Pseudoalteromonas spp., which peaked a b 250 ) 4 when linear and simple aromatic hydro- 2 carbons were abundant (Dubinsky 200 500 et al., 2013; Kleindienst et al., 2015a). Employing metagenomics, scientists Seawater (x10 Seawater

–1 150 confirmed that the different hydrocarbon 1,000 1 degradation genes and pathways present Depth (m) A B B A A 100 at different time points corresponded with abundances of different substrates 1,500

Copies ml pmoA Copies within the plume (Mason et al., 2012; 50 0 20, 2010) Explosion (April Since Days Redmond and Valentine, 2012). Meanwhile, in beach sands, succes- May 2,000 2 3 4 Pre-Spill May/June Aug/Sep Nov/Dec 10 10 10 sion of microbial populations also par- pmoA Copies ml–1 Seawater alleled the chemical evolution of petro-

FIGURE 2. Abundance of pmoA genes following the Deepwater Horizon discharge. (a) Abundance leum hydrocarbons (Rodriguez-R et al., of methanotrophic bacterial pmoA genes. Data are binned by time period. (b) Abundance of pmoA 2015). Early responders were mostly gene copies over time (pre-spill samples are magenta). Stars = OPU1. Diamonds = OPU3. Squares = Gammaproteobacteria (Alcanivorax and New phenotype. Data in the two time periods marked with asterisks in panel (a) are different from those collected at other times, with a statistical significance of a P value of <0.05. Plus signs denote Marinobacter spp.), microbes known extreme data outliers. Figure from Crespo-Medina et al. (2014) to degrade aliphatic hydrocarbons, and

100 Gammaproteobacteria; Oceanospirillales 90 Other 80 Gammaproteobacteria; Colwellia 70 Crenarchaeota; Nitrosopumilus 60 Gammaproteobacteria; Cycloclasticus

50 Bacteria; Marine Group A Gammaproteobacteria; Thalassomonas 40 Proteobacteria; Deltaproteobacteria 30 Relative Abundance (%) Abundance Relative Proteobacteria; Gammaproteobacteria 20 Euryarchaeota; Marine Group II 10 Alphaproteobacteria; Candidatus Pelagibacter 0 Uncontaminated Proximal Plume Distal Plume Sample Pyrotag Pyrotag Pyrotag

FIGURE 3. Microbial community structure and function change drastically in response to hydrocarbons. Community structure, function, and succession patterns varied depending on location and time of sampling. Figure modified from Mason et al. (2012)

128 Oceanography | Vol.34, No.1 their bloom coincided with a drastic In anoxic environments (e.g., the sea- Researchers studying the DWH spill decline in community diversity (Kostka floor and salt marsh sediments), biodeg- discovered that nitrogen fixation is a key et al., 2011). These were replaced after radation occurs much more slowly so that ecosystem function of microbial com- three months by Alphaproteobacteria most of the oil persists. (Liu et al., 2012; munities in response to oil (Rodriguez-R (Hyphomonas and Parvibaculum spp.), Chanton et al., 2015). Yet, seafloor sedi- et al., 2015; Fernández-Carrera et al., capable of aromatic hydrocarbon decom- ments near the DWH wellhead showed 2016). In the offshore and coastal ocean position. This shift coincided with the dis- some enrichment, notably in anaerobic ecosystems where nitrogen is limit- appearance of alkanes and persistence of respiration and anaerobic hydrocarbon ing, nitrogen fixation function increases aromatics in field samples (Huettel et al., degradation genes and associated species in response to oil contamination (Scott 2018). Nearly all of the oil on Pensacola such as Deltaproteobacteria (Mason et al., et al., 2014; Rodriguez-R et al., 2015; Beach was degraded after one year, and 2014b). Novel anaerobic polycyclic aro- Gaby et al., 2018; Shin et al., 2019a), and microbial communities returned to a matic hydrocarbon (PAH) degradation then dissipates once the hydrocarbons are state that resembled the pre-oil condi- pathways were also discovered (Shin et al., significantly reduced (Shin et al., 2019a). tion, with nearly all known hydrocarbon 2019b), potentially mediated by yet undis- Fixing nitrogen provides a strong selec- degraders and genes coding for degra- covered sulfate-reducing bacteria along tive advantage for those species capable dation functions exhibiting only low or with their fermentative syntrophic part- of fixation in the presence of oil, allow- undetectable levels (Rodriguez-R et al., ners. Hexadecane-degrading microbes, ing them to utilize hydrocarbon energy 2015). However, significant differences closely related to the Desulfobacteraceae, sources without restriction. Although were observed between the pre-oiled and phenanthrene-degrading microbes species with such dual capabilities are and recovered communities. Taxonomic related to Desulfatiglans spp. were uncommon under pristine conditions, diversity remained elevated in recovered also discovered. they thrive in oil-contaminated envi- communities, above that of the pre-oiled ronments. This explains the niche dom- community, while functional diversity Biogeochemical Processes inance of keystone oil-degrading organ- was lower in the recovered compared to Microbes play fundamental roles in basic isms such as Candidatus Macondimonas the pre-oiled. Such differences could be biogeochemical cycles of marine and diazotrophica (Karthikeyan et al., 2019). due to unrecognized long-term effects coastal ecosystems, including intercon- Microbial genes involved in car- of disturbance, such as the emergence nected carbon, nitrogen, and phosphorus bon, nitrogen, phosphorus, sulfur, and of new taxa, and stochastic environmen- cycling—and also in hydrocarbon degra- iron cycling were also enriched in oil-​ tal events such as changes in nutrient or dation. Omics analyses of these pathways contaminated ecosystems (for additional carbon input (Kostka et al., 2020). This revealed how these ecosystem functions information, see Mason et al., 2012; return to a “baseline” state remains an were impacted by the DWH spill. Rodriguez-R et al., 2015). Understanding open question that is actively being pur- these biogeochemical interdependencies sued as a legacy DWH issue. Nitrogen Fixation could inform fertilization strategies to In salt marshes, the microbial diver- Like all microbes, oil degraders require enhance natural biodegradation. sity of aerobic and anaerobic sediments substantial amounts of nitrogen for decreased, favoring microbes capable growth, but nitrogen is present only in Marine Oil Snow (MOS) of degrading alkanes and aromatics as limited amounts in marine environments. Phytoplankton and hydrocarbon-degrad​ - weathered oil and residues accumulated Although nitrogen fertilizers have been ing microbes that assemble around oil (Atlas et al., 2015). In anoxic sediments, added to accelerate microbial growth and droplets in large communities produce sulfate-reducing bacteria, for example, biodegradation following some oil spills, transparent exopolymers (TEPs) that Desulfococcus spp., increased in paral- such as the Exxon Valdez spill in Prince entrap oil. These dense macroscopic lel with oil contamination. Interestingly, William Sound, Alaska (Pritchard et al., assemblages of carbohydrates and bio- marsh sediments maintained a high per- 1992), microbes can generate their own mass behave like “snow” as they sink to centage of hydrocarbon degraders whose fixed nitrogen through biochemical fix- the seafloor. Marine oil snow (MOS) growth was driven periodically by storm ation of dinitrogen gas (Foght, 2010). effectively transfers hydrocarbons from surges that redistributed weathered oil Nitrogen fixation pathways are well the water column to the seafloor as sed- from the anoxic sediment. These lin- known in soil microbes that support crop iment (Vonk et al., 2015). Although MOS gering shifts in microbial community growth and in oceanic microbes such as formation was observed during previ- diversity and function have the poten- Trichodesmium cyanobacteria (Bergman ous oil spills, including the 1977 Baltic tial to impact the overall marsh eco- et al., 2013), but nitrogen fixation by oil Sea Tsesis spill (Johansson et al., 1980) system, particularly marsh vegetation degraders in response to raw oil exposure and the 1979 Ixtoc I spill in the south- and fishery health. is a new discovery (Quigg et al., 2016). ern Gulf of Mexico (Patton et al., 1981),

Oceanography | March 2021 129 it was first thoroughly studied after the tus, fecal matter, minerals, and chemical SIGNIFICANCE AND DWH spill, which triggered MOS for- dispersants, as well as such living cells as APPLICATIONS OF GoMRI mation in unprecedented quantities diatoms, microalgae, and bacteria, and RESEARCH (Joye et al., 2014). the polymers they secrete. Ultimately, scientific and technological Oil degradation and deposition that The question remains, however, advances derived from GoMRI research takes places in MOS impacts the fate of whether MOS is good or bad as an oil provide a foundation for new research spilled oil (Figure 4). MOS particles are spill bioremediation mechanism. MOS that will improve strategies for predict- hot spots for oil degradation, exhibiting could have negative consequences for ing and mitigating ecosystem perturba- high levels of lipase activity (Gutierrez the water column as well as the seafloor tions and have impacts across broad sci- et al., 2018) and distinct microbial where it is ultimately deposited. These entific disciplines and far beyond the communities that specialize in break- larger particles can trap and remove Gulf of Mexico. ing down oil. In particular, Colwellia, planktonic animals from the water col- Marinobacter spp., and Alteromonas spp. umn. MOS particles can also effectively Broad Lessons in Microbial are prevalent because they have the seal off the sediment surface, poten- Hydrocarbon Response unique capacity to degrade oil relatively tially suffocating fauna, while addition As the examples cited above show, rapidly in cold, deep marine environ- of organic content to the benthos could GoMRI research has revealed ecological, ments (Gutierrez et al., 2018). stimulate respiration, with negative evolutionary, and biogeochemical princi- MOS particles sink quickly, reaching consequences for filter-feeding fauna. ples that underlie microbial community the seafloor within ~10 days (Passow, Understanding the consequences of MOS composition and response to disturbance 2016). MOS formation is triggered by in the water column and the benthos across diverse environments and contexts. water column elements that include detri- requires further investigation. The hydrocarbon-degrading microbes and genes that code for hydrocarbon degradation and utilization discovered by GoMRI scientists are present in oil- contaminated waters and coastal regions Oil layer and droplets around the world, and thus are available everywhere as first responders. However, of phytoplankton, detritus, while oil-degrading microbes are globally and minerals, including oil bacterial mucus production, zooplankton feces, distributed, regional variables can signifi- and feeding webs with oil cantly influence microbial community composition and metabolic capabilities prior to and during a spill. Physical, chemical, and biological envi- ronmental conditions, including tem- Microbial modi cation perature, salinity, hydrostatic pressure, and degradation oxygen supply, nutrient availability (par- Zooplankton interaction —grazing and repackaging ticularly N and P), water currents and stratification, MOS presence, and sea Incorporation of ice can influence microbial species com- lithogenic material position and rates of hydrocarbon deg- radation, activity, and dispersal from the deep sea to intertidal marsh ecosys- tems (Head et al., 2006; Edwards et al., 2011; Redmond and Valentine, 2012; Rodriguez-R et al., 2015; Fernández- Benthic fauna Carrera et al., 2016; Huettel et al., 2018; Sun and Kostka, 2019). Furthermore, the type of oil spilled, how it weathers, and the extent of sunlight exposure can alter FIGURE 4. The formation, sinking and loss of marine oil snow, which consists of exudates, biotic, the chemical composition of oil and/or and abiotic particles. Marine oil snow is a common food source that removes hydrocarbons from the water column as it sinks to the seafloor. Image credit: Uta Passow, Memorial University induce MOS formation, all of which act of Newfoundland to drive different microbial responses

130 Oceanography | Vol.34, No.1 (Bacosa et al., 2015; Shin et al., 2019a). Informing Oil Spill Response traffic (Sun and Kostka, 2019). After a The DWH spill provided a unique oppor- GoMRI research provides a foundation of spill, monitoring mitigation and resto- tunity to study many of these environ- knowledge and tools that will enable sci- ration efforts should be undertaken to mental factors because it affected a wide entists to play key roles in guiding future assess whether response actions work as variety of open-water and coastal ecosys- oil spill response and mitigation planning predicted or if alternative plans should tems with large variations in physical and using data-driven strategies. Evidence- be implemented. chemical conditions. based interventions informed by evalu- In summary, with application of omics Microbial communities in the Gulf ating ecosystem responses and recovery tools, scientists can assist first responders of Mexico are considered to have been status will improve the success of future in determining potential environmental uniquely primed for rapid response to emergency responses by replacing previ- impacts of a spill, as well as actions to be oil spills due to the prevalence of natu- ous trial-and-error approaches. taken, on what time frames, and in which ral hydrocarbon seeps and drilling oper- Specifically, by examining the pres- locations to minimize risks and damage. ations that constantly discharge oil into ence of oil-degrading species, genes, and the Gulf. The background levels of hydro- biogeochemical pathways, scientists can Assessing Ecosystem Health in carbons sustain a higher proportion of assess the natural bioremediation poten- the Context of Environmental oil-degrading microbes that are also typ- tial of a community, providing insight Disturbance ical of many other regions with natu- into whether and how its metabolic capa- By developing basic principles of micro- ral seepage and/or hydrothermal vent- bilities can be augmented. One exam- bial ecosystem function and response ing, for example, the Gulf of California, ple is addition of nitrogen fertilizers, along with techniques to study them, the Bay of Bengal, the Black Sea, and the which has been proposed by GoMRI and GoMRI scientists have provided a basis Arctic Ocean. In addition, shipping lanes other investigators to accelerate micro- for understanding diverse marine and in open waters and ports are primed for bial bioremediation (Edwards et al., 2011; terrestrial ecosystems in order to address oil response because of long-term indus- Kleindienst et al., 2015b; Fernández- natural and man-made disasters, from trial and human activities. In some cases, Carrera et al., 2016). Additionally, chemical spills to algal blooms, and even hydrocarbons produced by phytoplank- enhancement of microbes that produce gradual and long-term stressors, includ- ton or cyanobacteria can prime microbes exopolymeric substances, which emul- ing climate change. Understanding for hydrocarbon degradation through sify oil and act like dispersants, could microbial ecosystem dynamics and low level exposure (Guiterrez et al., 2014; preclude addition of man-made disper- function in response to climate pertur- Love et al., 2021). sants such as Corexit (Ziervogel et al., bations will be essential for predicting Oil spills are particularly threatening 2019). Strategies to promote growth of long-term impacts and for preparing to Arctic regions where climate change native microbial species or utilization of mitigation plans. is already wreaking havoc. It is there- genetically engineered microbes for nat- Documentation of indigenous micro- fore critical to understand the environ- ural surfactant production would provide bial communities in diverse marine mental risks and develop mitigation responders with biodegradable, nontoxic and terrestrial ecosystems around the strategies before a major spill occurs in surfactant alternatives. Natural micro- world is necessary for assessing ecosys- these delicate ecosystems. Interestingly, bial surfactants may also prove effective tem changes and restoration goals in the GoMRI research has proven pertinent to as nontoxic dispersants with fewer envi- context of any perturbation. Baseline this region; preliminary results of these ronmental concerns (Bacosa et al., 2018). metagenomic analyses require sampling studies reveal microbial responses and Mounting an effective microbial oil over vast timescales to account for daily, community succession similar to those spill response will require analysis of monthly, and seasonal community vari- reported for the Arctic marine environ- microbial communities not only during a ations, as well as interannual and even ment (Sun and Kostka, 2019). However, spill but also before and after. Document- decadal shifts. Such studies are under- sea ice and extremely low temperatures ing the “normal” or “baseline” state of an way to catalogue Earth’s microbiome can alter degradation properties of both ecosystem prior to a crisis is essential for (Gilbert et al., 2018). From deep­water oil and microbial communities, which planning intelligent mitigation strategies benthic regions to coral reefs, beach may significantly impact biodegradation aimed at restoring native microbial com- sands, marshes, and associated terres- and need to be considered. The Canadian munities and their ecosystem functions trial ecosystems (Magnuson, 1990, 1995), government is actively engaged with the (Joye, 2015). In addition, it can indicate different ecosystems each have unique GoMRI community to build on GoMRI whether an ecosystem is compromised by microbial baselines that require individ- discoveries, methodology, and technol- previous environmental stress and there- ual documentation so that significant ogy in efforts to prepare for a spill in the fore in need of additional protections, for or problematic changes can be detected Canadian Arctic. example, limiting oil drilling or tanker and measured.

Oceanography | March 2021 131 Microbial Indicators of Ecosystem These genetic and metabolic indicators emergence of novel genes, metabolic Health and Disturbance of ecosystem disruption can also provide pathways, and ecosystem function can GoMRI researchers identified micro- critical clues as to how the environment all assist scientists in identifying biogeo- bial indicators of ecosystem disturbance has changed, and therefore how it might chemical disruptions. in the context of the DWH spill that will be restored. help scientists assess ecosystem health Therapeutic biomarkers. Indicator spe- across various contexts. THE FUTURE OF GENOMICS IN cies and genes might also point scientists THE STUDY AND MANAGEMENT and managers toward solutions by pro- Low microbial diversity can be a red OF ECOSYSTEM HEALTH viding information about how an envi- flag, signaling ecosystem disruption and In the future, scientists hope to identify ronment has changed and, therefore, declining ocean health. A healthy micro- microbial “biomarkers” of ocean health, how its ecosystem might be restored. bial ecosystem is composed of a variety as detectors of ecosystem disruption Therefore, in addition to informing mit- of species that provide diverse ecosys- and risk predictors in the face of diverse igation and restoration strategies, these tem functions and maintain system sta- threats, including man-made and natu- indicators can help to identify action bility. Environmental disturbances such ral disasters as well as climate change. For plans during a crisis event. For example, as the DWH spill disrupt the balance, example, how might microbiologists help the emergence of a nitrogen-fixing spe- favoring a limited number of species that fisheries or oyster farms recover after an cies might indicate nitrogen as a limiting is able to adapt to new conditions. These oil spill or hurricane? Similar to ways that factor in bioremediation, suggesting that few species thrive over the others, often medical biomarkers help doctors diagnose the addition of nitrogen fertilizers would reflected in low overall species diversity. diseases and identify appropriate treat- accelerate recovery. A community with low diversity may be ments, researchers could “take the pulse” more sensitive and less resilient to further of the ecosystem with microbial biomark- To assemble a complete picture of eco- disturbances. ers to identify atypical functioning and system health will require the use of multi- direct strategies to correct these problems. ple, diverse biomarkers, each of which will The presence of a single species or Biomarkers might be used in vari- provide a piece of the puzzle. For exam- genus in very high numbers indicates ous ways to assess and support ecosys- ple, genomic biomarkers may provide a recent ecosystem disruption, even if tem health: insight into hydrocarbon degradation, the original composition of the commu- toxin accumulation, hypoxic conditions, nity is unknown. For example, during Predictive biomarkers. Disruptions in or other environmental disturbances. the DWH spill, rare species with hydro- microbial communities can often be per- Meanwhile, metatranscriptomics might carbon degradation potential and/ ceived before ecological consequences reveal immediate adaptations, whereas or nitrogen fixation capabilities were or impacts are apparent. Thus, microbes meta-community restructuring suggests poised to take advantage of new energy can serve as the “canary in the coal mine” long-term consequences. sources. Under novel conditions, these to warn of impending consequences of Ideally, such biomarkers will be able to species outcompete their neighbors a disturbance. For example, microbial inform scientists and first responders on and can grow to one-third or more of community restructuring or functional site and in real time as an essential com- the population. changes might indicate a slow oil leak or ponent of an environmental emergency gradual warming due to climate change. response toolkit. This will require robust, The emergence of different and novel This can trigger early action to better pre- portable (i.e., hand-held devices), cost- genes, metabolic pathways, and eco- vent or mitigate potential damages from effective, and integrative analytic instru- system functions within microbial com- natural and man-made disasters in the ments, such as the small mobile DNA munities reflects adaptation to environ- ocean, as well as other ecosystems, by sequencers designed for the rigors of mental change. When microbes respond alerting scientists to changes before they field research. to environmental disturbance, organ- are irreversible, and/or identifying areas isms may express different genes and that need extra protections. Predicting Environmental Impacts metabolic pathways as they adapt to new with Biomarkers and Models sources of energy (hydrocarbons), nutri- Diagnostic biomarkers. The presence of Biomarkers can be paired with biogeo- ents, physical/chemical limitations, and specific indicator species and genes might chemical models to predict import- other factors in order to take on novel be used to “diagnose” certain distur- ant outcomes, including when oil will and adaptive ecosystem functions. bances, which could indicate what has be removed from a system, how long gone awry within an ecosystem. For would it take for an ecosystem to recover, example, overgrowth of a species and whether responders should intervene

132 Oceanography | Vol.34, No.1 to speed ecosystem recovery, and other lating omics data to metabolic func- resource can be mined to predict conse- aspects of biogeochemical responses and tions. While metagenomics informs quences and inform response. timelines. Models can also reveal which metabolic potential, it may not always Specifically, GoMRI researchers created factors may have the greatest impact on reflect microbial activity. There are many the Genome Repository of Oil Systems damage or recovery, enabling responders novel microbial species, genes, and pro- (Karthikeyan et al., 2020b), a comprehen- to tailor mitigation strategies rather than teins to be discovered and identified, and sive, searchable database that documents relying on one-size-fits-all approaches their biogeochemical functions charac- microbial populations in natural oil eco- that may or may not work. terized, most probably with inferences systems and oil spills, along with avail- Biogeochemical models can help sci- from taxonomic and sequence similari- able underlying physicochemical data, entists answer questions, including: ties to reference species and their meta- geocoded via the geographic information • Should oil spills be seeded with bolic profiles. In reality, metabolic capa- system to reveal their distribution pat- microbes with hydrocarbon degrada- bilities can differ dramatically from terns. Provided as an independent proj- tion capabilities? If so, which ones and predicted activities and currently must be ect through the Microbial Genomes Atlas in what quantity? defined through cultivation, biochemi- (MiGA) web server (http://microbial-​ • Should dispersants and/or fertilizers cal assays, or gene knockout experiments. genomes.org/), the repository contains be added? If so, in what quantity? How Machine learning/artificial intelligence over ~2,000 genomes, more than 95% of will such interventions impact ability of is likely to be of great value in over- which represent novel taxa, though rep- indigenous microbes to degrade oil? coming this obstacle. resentation of cultured organisms from • What environmental impact can be To better link omics data with func- oil-contaminated and oil reservoir eco- avoided with an intervention? What tional rates of activity at the enzymatic systems in this database is limited. The are potential unintended consequences and cellular levels will require new strate- database allows researchers to classify of intervention? gies, such as stable isotope probing (SIP) unknown genomes by reference to known • Will leveraging natural microbial pro- and biorthogonal noncanonical amino genomes, thereby facilitating the pre- cesses be more successful than alter- acid tagging (BONCAT). These tools dictive understanding of microbial taxa native approaches such as burning or track metabolic activity of individual and activities that can control the fate vacuuming the oil? microbes and consortia within their nat- of oil spills.

Using the powerful and constantly evolving tools of genomics, “ GoMRI researchers discovered novel [microbial] genes, pathways, organisms, communities, and partnerships associated with oil decomposition and spill remediation.”

Current biogeochemical models are ural environments. They can assign spe- CONCLUSIONS limited and, in general, unable to answer cific paths to specific microbes, elucidate Microbiology and omics tools have proven such complex questions. To achieve cross-feeding partnerships, and identify instrumental in providing an under- more realistic models will require signif- novel metabolic steps and pathways. standing of the impacts of the Deepwater icant advances in understanding phys- Horizon oil spill on marine and coastal ical, chemical, and biological processes Facilitating Advances Through ecosystems. The new omics-based​ tech- and their interactions in the ocean, as Open-Access Resources and nologies and strategies enabled a compre- well as active and effective collaborations Multidisciplinary Collaboration hensive investigation of Gulf microbial among oceanographers, microbiologists, The omics and environmental data sets communities and their biogeochemical computational scientists, and many collected by GoMRI investigators are functions in unprecedented detail. other disciplines. accessible via the open-access platform Using the powerful and constantly The greatest hurdle in developing pre- https://data.gulfresearchinitiative.org/. evolving tools of genomics, GoMRI dictive biogeochemical models is trans- In the event of a new spill or disaster, this researchers discovered novel genes, path-

Oceanography | March 2021 133 ways, organisms, communities, and part- phy following a deepwater oil-well blowout. Nature Huettel, M., W.A. Overholt, J.E. Kostka, C. Hagan, Geoscience 7:423–427, https://doi.org/10.1038/ J. Kaba, W.B. Wells, and S. Dudley. 2018. nerships associated with oil decomposi- NGEO2156. Degradation of Deepwater Horizon oil buried in a tion and spill remediation. Key shifts in Doyle, S.M., E.A. Whitaker, V. De Pascuale, T.L. Wade, Florida beach influenced by tidal pumping. Marine A.H. Knap, P.H. Santchi, A. Quigg, and J.B. Sylvan. Pollution Bulletin 126:488–500, https://doi.org/​ community structures that dictate essen- 2018. Rapid formation of microbe-oil aggregates 10.1016/j.marpolbul.2017.10.061. tial ecosystem functions and bioreme- and changes in community composition in coastal Johansson, S., U. Larsson, and P. Boehm. 1980. surface water following exposure to oil and the dis- The Tsesis oil spill impact on the pelagic eco- diation services were documented. persant Corexit. Frontiers in Microbiology 9:689, system. Marine Pollution Bulletin 11:284–293, Altogether, GoMRI studies have revealed https://doi.org/10.3389/fmicb.2018.00689. https://doi.org/​10.1016/0025-326X(80)90166-6. Doyle, S.M., G. Lin, M. Morales-McDevitt, T.L. Wade, Joye, S.B., I. Leifer, I.R. MacDonald, J.P. Chanton, new insights and core lessons about A. Quigg, and J.B. Sylvan. 2020. Niche partitioning C.D. Meile, A.P. Teske, J.E. Kostka, L. Chistoserdova, microbial community responses to envi- between coastal and offshore shelf waters results R. Coffin, D. Hollander, and M. Kastner. 2011. in differential expression of alkane and polycy- Comment on “A persistent oxygen anomaly reveals ronmental disturbances, as well as their clic aromatic hydrocarbon catabolic pathways. the fate of spilled methane in the deep Gulf of roles in maintenance and restoration of mSystems 5(4):e00668-20, https://doi.org/10.1128/ Mexico.” Science 332:1,033, https://doi.org/10.1126/ mSystems.00668-20. science.1203307. ecosystem stability during and after an Dubinsky, E.A., M.E. Conrad, R. Chakraborty, Joye, S.B., A.P. Teske, and J.E. Kostka. 2014. oil spill. The lessons learned inform a sci- M. Bill, S.E. Borglin, J.T. Hollibaugh, O.U. Mason, Microbial dynamics following the Macondo oil M.Y. Piceno, F.C. Reid, W.T. Stringfellow, and others. well blowout across Gulf of Mexico environments. entific understanding of how oil spills 2013. Succession of hydrocarbon-degrading bac- Bioscience 64:766–777, https://doi.org/10.1093/ impact various ecosystems, from deep teria in the aftermath of the Deepwater Horizon oil biosci/biu121. spill in the Gulf of Mexico. Environmental Science Joye, S.B. 2015. Deepwater Horizon, 5 years on. waters to beach sands to tidal marshes, & Technology 47:10,860–10,867, https://doi.org/​ Science 349(6248):592–593, https://doi.org/​ and will allow improved prediction 10.1021/es401676y. 10.1126/​science.aab4133. Edwards, B.R., C.M. Reddy, R. Camilli, Joye, S.B., S. Kleindienst, J.A. Gilbert, and mitigation of damage in the event C.A. Carmichael, K. Longnecker, and K.M. Handley, P. Weisenhorn, W.A. Overholt, of future spills. B.A.S. Van Mooy. 2011. Rapid microbial respira- and J.E. Kostka. 2016. Responses of micro- tion of oil from the Deepwater Horizon spill in bial communities to hydrocarbon exposures. offshore surface waters of the Gulf of Mexico. Oceanography 29(3):136–149, https://doi.org/​ RESOURCES Environmental Research Letters 6:035301, 10.5670/oceanog.2016.78. • http://www.taraoceans-dataportal.org/ https://doi.org/10.1088/1748-9326/6/3/035301. Kamalanathan, M., S.M. Doyle, C. Xu, A.M. Achberger, • https://data.gulfresearchinitiative.org/ Fernández-Carrera, A., K.L. Rogers, S.C. Weber, T.L. Wade, K. Schwehr, P.H. Santschi, J.B. Sylvan, J.P. Chanton, and J.P. Montoya. 2016. Deep Water and A. Quigg. 2020. Exoenzymes as a signature of • http://www.mg-rast.org/ Horizon oil and methane carbon entered the microbial response to marine environmental con- • http://enve-omics.ce.gatech.edu/ food web in the Gulf of Mexico. Limnology and ditions. mSystems 5(2):e00290-20, https://doi.org/​ Oceanography 61:S387–S400, https://doi.org/​ 10.1128/mSystems.00290-20. REFERENCES 10.1002/lno.10440. Karthikeyan, S., L.M. Rodriguez-R, P. Heritier- Atlas, R.M., D.M. Stoeckel, S.A. Faith, A. Minard- Foght, J. 2010. Nitrogen fixation and hydrocarbon-​ Robbins, M. Kim, W.A. Overholt, J.C. Gaby, J.K. Hatt, Smith, J.R. Thorn, and M.J. Benotti. 2015. Oil bio- oxidizing bacteria. Pp. 431–448 in Cellular J.C. Spain, R. Rosselló-Móra, M. Huettel, and degradation and oil-degrading microbial popula- Ecophysiology of Microbe: Hydrocarbon and others. 2019. “Candidatus Macondimonas diazotro- tions in marsh sediments impacted by oil from the Lipid Interactions. Handbook of Hydrocarbon and phica,” a novel gammaproteobacterial genus dom- Deepwater Horizon well blowout. Environmental Lipid Microbiology. T. Krell, ed., Springer, Cham, inating crude-oil-contaminated coastal sediments. Science & Technology 49:8,356–8,366, https://doi.org/10.1007/978-3-319-50542-8_53. The ISME Journal 13:2,129–2,134, https://doi.org/​ https://doi.org/​10.1021/acs.est.5b00413. Gaby, J.C., L. Rishishwar, L.C. Valderrama-Aguirre, 10.1038/s41396-019-0400-5. Bacosa, H.P., D.L. Erdner, and Z. Liu. 2015. S.J. Green, A. Valderrama-Aguirre, I. King Jordan, Karthikeyan, S., M. Kim, P. Heritier-Robbins, Differentiating the roles of photooxidation and J.E. Kostka. 2018. Diazotroph community char- J.K. Hatt, J.C. Spain, W.A. Overholt, M. Huettel, and biodegradation in the weathering of Light acterization via a high-throughput nifH ampli- J.E. Kostka, and K.T. Konstantinidis. 2020a. Louisiana sweet crude oil in surface water from con sequencing and analysis pipeline. Applied Integrated omics elucidate the mechanisms driv- the Deepwater Horizon site. Marine Pollution and Environmental Microbiology 84:e01512-17, ing the rapid biodegradation of Deepwater Bulletin 95:265–272, https://doi.org/10.1016/​ https://doi.org/10.1128/AEM.01512-17. Horizon oil in intertidal sediments undergoing j.marpolbul.2015.04.005. Gilbert, J.A., J.K. Jansson, and R. Knight. 2018. Earth oxic-anoxic cycles. Environmental Science & Bacosa, H.P., D.L. Erdner, B.E. Rosenheim, microbiome project and global systems biology. Technology 54(16):10,088–10,099, https://doi.org/​ P. Shetty, K.W. Seitz, B.J. Baker, and Z. Liu. 2018. mSystems 3(3):e00217-17, https://doi.org/10.1128/ 10.1021/acs.est.0c02834. Hydrocarbon degradation and response of sea- mSystems.00217-17. Karthikeyan, S., L.M. Rodriguez-R, P. Heritier- floor sediment bacterial community in the northern Grossart, H.P., R. Massana, K.D. McMahon, and Robbins, J.K. Hatt, M. Huettel, J.E. Kostka, and Gulf of Mexico to light Louisiana sweet crude oil. D.A. Walsh. 2020. Linking metagenomics to K.T. Konstantinidis. 2020b. Genome repository of The ISME Journal 12:2,532–2,543, https://doi.org/​ aquatic microbial ecology and biogeochemical oil systems: An interactive and searchable data- 10.1038/s41396-018-0190-1. cycles. Limnology and Oceanography 65:S2–S20, base that expands the catalogued diversity of Bergman, B., G. Sandh, S. Lin, J. Larsson, and https://doi.org/10.1002/lno.11382. crude oil-associated microbes. Environmental E.J. Carpenter. 2013. Trichodesmium – A wide- Gutierrez, T., G. Morris, D. Ellis, B. Bowler, Microbiology 22(6):2,094–2,106, https://doi.org/​ spread marine cyanobacterium with unusual M. Jones, K. Salek, B. Mulloy, and A. Teske. 2018. 10.1111/1462-2920.14966. nitrogen fixation properties. FEMS Microbiology Hydrocarbon-degradation and MOS-formation Kostka, J.E., O. Prakash, W.A. Overholt, S.J. Green, Reviews 37:286–302, https://doi.org/​10.1111/​ capabilities of the dominant bacteria enriched in G. Freyer, A. Canion, J. Delgardio, N. Norton, j.1574-6976.2012.00352.x. sea surface oil slicks during the Deepwater Horizon T.C. Hazen, and M. Huettel. 2011. Hydrocarbon- Chanton, J., T. Zhao, B.E. Rosenheim, S. Joye, oil spill. Marine Pollution Bulletin 135:205–215, degrading bacteria and the bacterial community S. Bosman, C. Brunner, K.M. Yeager, A.R. Diercks, https://doi.org/10.1016/j.marpolbul.2018.07.027. response in Gulf of Mexico beach sands impacted and D. Hollander. 2015. Using natural abundance Hazen, T.C., E.A. Dubinsky, T.Z. DeSantis, by the Deepwater Horizon oil spill. Applied and radiocarbon to trace the flux of petrocarbon to the G.L. Andersen, Y.M. Piceno, N. Singh, J.K. Jansson, Environmental Microbiology 77:7,962–7,974, seafloor following the Deepwater Horizon oil spill. A. Probst, S.E. Borglin, J.L. Fortney, and others. https://doi.org/10.1128/AEM.05402-11. Environmental Science & Technology 49:847–854, 2010. Deep-sea oil plume enriches indigenous Kostka, J.E., W.A. Overholt, L.M. Rodriguez-R, https://doi.org/10.1021/es5046524. oil-degrading bacteria. Science 330:204–208, M. Huettel, and K. Konstantinidis. 2020. Toward a Crespo-Medina, M., C.D. Meile, K.S. Hunter, https://doi.org/10.1126/science.1195979. predictive understanding of the benthic microbial A.-R. Diercks, V.L. Asper, V.J. Orphan, Head, I.M., D.M. Jones, and W.F.M. Roling. 2006. community response to oiling on the northern Gulf P.L. Tavormina, L.M. Nigro, J.J. Battles, J.P. Chanton, Marine microorganisms make a meal of oil. Nature of Mexico coast. Pp. 182–202 in Scenarios and and others. 2014. The rise and fall of methanotro- Reviews Microbiology 4:173–182, https://doi.org/​ Responses to Future Deep Oil Spills. S. Murawski, 10.1038/nrmicro1348.

134 Oceanography | Vol.34, No.1 C.H. Ainsworth, S. Gilbert, D.J. Hollander, C.B. Paris, Patton, J.S., M.W. Rigler, P.D. Boehm, and D.L. Fiest. Zengler, K., and L.S. Zaramela. 2018. The social net- M. Schlüter, and D.L. Wetzel, eds, Springer, 1981. Ixtoc I oil spill: Flaking of surface mousse work of microorganisms—how auxotrophies https://doi.org/10.1007/978-3-030-12963-7_11. in the Gulf of Mexico. Nature 290:235–238, shape complex communities. Nature Reviews Kleindienst, S., S. Grim, M. Sogin, A. Bracco, https://doi.org/​10.1038/290235a0. Microbiology 16(6):383–390, https://doi.org/10.1038/ M. Crespo-Medina, and S.B. Joye. 2015a. Diverse, Pritchard, P.H., J.G. Mueller, J.C. Rogers, s41579-018-0004-5. rare microbial taxa responded to the Deepwater F.V. Kremer, and J.A. Glaser. 1992. Oil spill Ziervogel, K., S.B. Joye, S. Kleindienst, S.Y. Malkin, Horizon deep-sea hydrocarbon plume. The ISME bioremediation: Experiences, lessons and U. Passow, A.D. Steen, and C. Arnosti. 2019. Journal 10:400–415, https://doi.org/10.1038/ results from the Exxon Valdez oil spill in Alaska. Polysaccharide hydrolysis in the presence of oil ismej.2015.121. Biodegradation 3:315–335, https://doi.org/​10.1007/ and dispersants: Insights into potential degrada- Kleindienst, S., M. Seidel, K. Ziervogel, S. Grim, BF00129091. tion pathways of exopolymeric substances (EPS) K. Loftis, S. Harrison, S.Y. Malkin, M.J. Perkins, Quigg, A., U. Passow, W.-C. Chin, C. Xu, S. Doyle, from oil-degrading bacteria. Elementa: Science J. Field, M.L. Sogin, and others. 2015b. Chemical L. Bretherton, M. Kamalanathan, A.K. Williams, of the Anthropocene 7:31, http://doi.org/10.1525/ dispersants can suppress the activity of natural J.B. Sylvan, Z.V. Finkel, and others. 2016. The elementa.371. oil-degrading microorganisms. Proceedings of the role of microbial exopolymers in determining the National Academy of Sciences of the United States fate of oil and chemical dispersants in the ocean. AUTHORS of America 112:14,900–14,905, https://doi.org/​ Limnology and Oceanography Letters 1:3–26, Shannon Weiman ([email protected]) is a free- 10.1073/pnas.1507380112. https://doi.org/10.1002/lol2.10030. lance science and biomedical writer/editor located Love, C.R., E.E. Arrington, K.M. Gosselin, C.M. Reddy, Redmond, M.C., and D.L. Valentine. 2012. Natural in Boulder, CO, USA. Samantha B. Joye is Professor, B.A.S. Van Mooy, R.K. Nelson, and D.L. Valentine. gas and temperature structured a micro- Department of Marine Science, University of Georgia, 2021. Microbial production and consumption bial community response to the Deepwater Athens, GA, USA. Joel E. Kostka is Professor, School of hydrocarbons in the global ocean. Nature Horizon oil spill. Proceedings of the National of Biology and Earth and Atmospheric Sciences, Microbiology 6:489–498, https://doi.org/10.1038/ Academy of Sciences of the United States of Georgia Institute of Technology, Atlanta, GA, USA. s41564-020-00859-8. America 109:20,292–20,297, https://doi.org/​ Kenneth M. Halanych is Professor, Department of Liu, Z., J. Liu, Q. Zhu, and W. Wu. 2012. The weath- 10.1073/​pnas.1108756108. Biological Sciences, Auburn University, Auburn, AL, ering of oil after the Deepwater Horizon oil spill: Rodriguez-R, L.M., W.A. Overholt, C. Hagan, USA. Rita R. Colwell is Distinguished University Insights from the chemical composition of the M. Huettel, J.E. Kostka, and K.T. Konstantinidis. Professor, University of Maryland, College Park, oil from the sea surface, salt marshes and sedi- 2015. Microbial community successional pat- MD, USA. ments. Environmental Research Letters 7:035302, terns in beach sands impacted by the Deepwater https://doi.org/10.1088/1748-9326/7/3/035302. Horizon oil spill. The ISME Journal 9:1,928–1,940, ARTICLE CITATION Liu, Z., and J. Liu. 2013. Evaluating bacterial com- https://doi.org/10.1038/ismej.2015.5. Weiman, S., S.B. Joye, J.E. Kostka, K.M. Halanych, munity structures in oil collected from the sea Scott, N.M., M. Hess, N.J. Bouskill, O.U. Mason, and R.R. Colwell. 2021. GoMRI insights into microbial surface and sediment in the northern Gulf of J.K. Jansson, and J.A. Gilbert. 2014. The microbial genomics and hydrocarbon bioremediation response Mexico after the Deepwater Horizon oil spill. nitrogen cycling potential is impacted by polyaro- in marine ecosystems. Oceanography 34(1):124–135, MicrobiologyOpen 2:492–504, https://doi.org/​ matic hydrocarbon pollution of marine sediments. https://doi.org/10.5670/oceanog.2021.121. 10.1002/mbo3.89. Frontiers in Microbiology 5:108, https://doi.org/​ Magnuson, J.J. 1990. Long-term ecologi- 10.3389/fmicb.2014.00108. cal research and the invisible present. Shin, B., I. Bociu, M. Kolton, M. Huettel, and COPYRIGHT & USAGE This is an open access article made available under BioScience 40(7):495–501, https://doi.org/​ J.E. Kostka. 2019a. Succession of microbial popula- the terms of the Creative Commons Attribution 4.0 10.2307/1311317. tions and nitrogen-fixation associated with the bio- International License (https://creativecommons.org/ Magnuson, J.J. 1995. The invisible present. degradation of sediment-oil-agglomerates buried licenses/by/4.0/), which permits use, sharing, adap- Pp. 448–464 in Ecological Time Series. in a Florida sandy beach. Scientific Reports 9:19401, tation, distribution, and reproduction in any medium T.M. Powell and J.H. Steele, eds, Springer, Boston, https://doi.org/10.1038/s41598-019-55625-6. or format as long as users cite the materials appro- MA, https://doi.org/10.1007/978-1-4615-1769-6_20. Shin, B., M. Kim, K. Zengler, K.J. Chin, W.A. Overholt, priately, provide a link to the Creative Commons Mason, O.U., T.C. Hazen, S. Borglin, P.S. Chain, L.M. Gieg, and K.T. Konstantinidis. 2019b. Anaerobic license, and indicate the changes that were made to E.A. Dubinsky, J.L. Fortney, J. Han, H.Y. Holman, degradation of hexadecane and phenanthrene the original content. J. Hultman, R. Lamendella, and others. 2012. coupled to sulfate reduction by enriched consor- Metagenome, metatranscriptome and single-​ tia from northern Gulf of Mexico seafloor sediment. cell sequencing reveal microbial response Scientific Reports 9:1239, https://doi.org/10.1038/ to Deepwater Horizon oil spill. The ISME s41598-018-36567-x. Journal 6:1,715–1,727, https://doi.org/10.1038/ Sun, X., and J.E. Kostka. 2019. Hydrocarbon- ismej.2012.59. degrading microbial communities are site-​ Mason, O.U., J. Han, T. Woyke, and J.K. Jansson. specific and their activity is limited by synergies 2014a. Single-cell genomics reveals features of in temperature and nutrient availability in sur- a Colwellia species that was dominant during face ocean waters. Applied and Environmental the Deepwater Horizon oil spill. Frontiers in Microbiology 85:e00443-19, https://doi.org/10.1128/ Microbiology 5:332, https://doi.org/10.3389/ AEM.00443-19. fmicb.2014.00332. Valentine, D.L., J.D. Kessler, M.C. Redmond, Mason, O.U., N.M. Scott, A. Gonzalez, A. Robbins- S.D. Mendes, M.B. Heintz, C. Farwell, L. Hu, Pianka, J. Baelum, J. Kimbrel, N.J. Bouskill, F.S. Kinnaman, S. Yvon-Lewis, M. Du, and others. E. Prestat, S. Borglin, D.C. Joyner, and others. 2010. Propane respiration jump-starts microbial 2014b. Metagenomics reveals sediment micro- response to a deep oil spill. Science 330:208–211, bial community response to Deepwater Horizon https://doi.org/10.1126/science.1196830. oil spill. The ISME Journal 8:1,464–1,475, Vonk, S.M., D.J. Hollander, and A.J. Murk. 2015. Was https://doi.org/​10.1038/ismej.2013.254. the extreme and wide-spread marine oil-snow sed- Newell, S.E., D. Eveillard, M.J. McCarthy, imentation and flocculent accumulation (MOSSFA) W.S. Gardner, Z. Liu, and B.B. Ward. 2014. A shift event during the Deepwater Horizon blow- in the archaeal nitrifier community in response out unique? Marine Pollution Bulletin 100:5–12, to natural and anthropogenic disturbances in https://doi.org/​10.1016/j.marpolbul.2015.08.023. the northern Gulf of Mexico. Environmental Yang, T., L.M. Nigro, T. Gutierrez, L. D’Ambrosio, Microbiology Reports 6:106–112, https://doi.org/​ S.B. Joye, R. Highsmith, and A. Teske. 2016. Pulsed 10.1111/1758-2229.12114. blooms and persistent oil-degrading bacterial pop- Passow, U. 2016. Formation of rapidly-sinking, oil-​ ulations in the water column during and after the associated marine snow. Deep Sea Research Deepwater Horizon blowout. Deep Sea Research Part II 129:232–240, https://doi.org/10.1016/​ Part II 129:282–291, http://doi.org/10.1016/j.dsr2.​ j.dsr2.2014.10.001. 2014.01.014.

Oceanography | March 2021 135