The cold waters and harsh environmental conditions in the Arctic have led some investigators to conclude that spilled oil will be more recalcitrant and remain in the environment for longer periods of time than oil spilled in less harsh environments. Other investigators have concluded that the microorganisms in the Arctic are adapted to those conditions and rapidly respond to spilled oil. The objective of this document is to examine experimental evidence from peer-reviewed literature that addresses these conclusions and identify areas of additional research. 

This chapter focuses on biodegradation of oil spilled into the marine Arctic environment. The state-of-knowledge will be summarized, and main uncertainties that need further addressing will be highlighted. In this summary we also describe some of the specific conditions that organisms involved in biodegradation processes encounter in the Arctic environment but not at lower latitudes. These differentiating conditions, which may affect biodegradation rates and capabilities in the Arctic, include generally low near-surface temperatures, ice coverage and (in open water) extreme fluctuations in the photoperiod during the Arctic summer and winter. 

The risk of oil pollution in the Arctic is relevant to both marine and terrestrial environments. The Arctic marine environment may be influenced by recently discovered oil and gas deposits, together with the potential for increased traffic attributable to shipping lanes made possible by historically low summer ice cover. The terrestrial environments may be exposed to pollution risk, for instance, from oil transport pipeline systems and other land-based activities. In this review we will focus on the marine environments, although references will also be made to terrestrial studies, when relevant. 

Oil compounds can be degraded by a variety of prokaryotic and eukaryotic organisms. After uptake in eukaryotic macro-organisms, provided they are at sub-lethal or sub-inhibitory concentrations, the compounds may be transformed by the metabolic apparatus, bio-accumulated, or removed through fecal excretion. Despite the range of organism types that can act on oil, however, the microbial communities are regarded as the main source for biological degradation of marine contamination. This perspective is based on the large concentrations of microorganisms, large surface-to-volume ratios and their rapid responses to the situation. This section will therefore focus on the microbial populations and their responses to possible oil discharges in the Arctic marine environment. 

Based on sequencing studies of the 16S rRNA gene, in the mid-1970s Carl Woese and colleagues separated living organisms into the domains Eukarya, Archaea and Bacteria (Woese et al. 1990). Microorganisms are present in all three domains. The bacteria and archaea are all microorganisms, though some (e.g. cyanobacteria) sometimes aggregate into visible mats or filaments. Among the eukaryotes are unicellular microorganisms often referred to as protists, which include the photosynthetic phytoplankton as well as some heterotrophic organisms (e.g. amoebae). Within this diverse collective of microorganisms, biodegradation of oil compounds is primarily associated with bacterial processes. In the marine environment the archaeal responses to oil pollution have mainly been considered low when compared to bacterial responses (e.g. Röling et al. 2004). However, archaea are important members of communities in petroleum reservoirs and are involved in the mineralization of petroleum hydrocarbons, for instance by methanogenesis under anoxic conditions. Recent studies have also implied the contribution of halophilic archaea to the biodegradation of hydrocarbons in the marine environment (Al-Mailem et al. 2010; Liu et al. 2009). Specific to the Arctic marine environment, essentially nothing is known of the archaeal contributions to oil degradation. 

Among eukaryotes, degradation of oil is mainly associated with fungi (Prince 2005). Only a few studies have been conducted to characterize fungal communities in Arctic marine environments (e.g. Gunde-Cimerman et al. 2003; Butinar et al. 2011). Among the microorganisms, our understanding of the protist contribution to oil biodegradation is also limited. Heterotrophic protists may affect petroleum biodegradation indirectly through their role as bacterial predators. Efficient grazing by protists can maintain the bacterial community in the log-growth phase, resulting in increased metabolic rates and thus increased biodegradation rates (Ballestero and Magdol 2011). 

In summary, most evidence for microbial degradation in the Arctic is connected to bacteria and bacterial communities. Our focus in this section will be on bacterial communities and the bacterial processes involved in oil biodegradation in the Arctic environment. However, the need for greater knowledge related to the contribution of archaeal and protistan organisms to oil biodegradation is an incentive to develop a better understanding of their contribution to oil biodegradation and microbial ecology in the Arctic. 

The cold waters and harsh environmental conditions in the Arctic have led some investigators to conclude that spilled oil will be more recalcitrant and remain in the environment for longer periods of time than oil spilled in less harsh environments. Other investigators have concluded that the microorganisms in the Arctic are adapted to those conditions and rapidly respond to spilled oil. The objective of this document is to examine experimental evidence from peer-reviewed literature that addresses these conclusions and identify areas of additional research. 

This chapter focuses on biodegradation of oil spilled into the marine Arctic environment. The state-of-knowledge will be summarized, and main uncertainties that need further addressing will be highlighted. In this summary we also describe some of the specific conditions that organisms involved in biodegradation processes encounter in the Arctic environment but not at lower latitudes. These differentiating conditions, which may affect biodegradation rates and capabilities in the Arctic, include generally low near-surface temperatures, ice coverage and (in open water) extreme fluctuations in the photoperiod during the Arctic summer and winter. 

The risk of oil pollution in the Arctic is relevant to both marine and terrestrial environments. The Arctic marine environment may be influenced by recently discovered oil and gas deposits, together with the potential for increased traffic attributable to shipping lanes made possible by historically low summer ice cover. The terrestrial environments may be exposed to pollution risk, for instance, from oil transport pipeline systems and other land-based activities. In this review we will focus on the marine environments, although references will also be made to terrestrial studies, when relevant. 

Oil compounds can be degraded by a variety of prokaryotic and eukaryotic organisms. After uptake in eukaryotic macro-organisms, provided they are at sub-lethal or sub-inhibitory concentrations, the compounds may be transformed by the metabolic apparatus, bio-accumulated, or removed through fecal excretion. Despite the range of organism types that can act on oil, however, the microbial communities are regarded as the main source for biological degradation of marine contamination. This perspective is based on the large concentrations of microorganisms, large surface-to-volume ratios and their rapid responses to the situation. This section will therefore focus on the microbial populations and their responses to possible oil discharges in the Arctic marine environment. 

Based on sequencing studies of the 16S rRNA gene, in the mid-1970s Carl Woese and colleagues separated living organisms into the domains Eukarya, Archaea and Bacteria (Woese et al. 1990). Microorganisms are present in all three domains. The bacteria and archaea are all microorganisms, though some (e.g. cyanobacteria) sometimes aggregate into visible mats or filaments. Among the eukaryotes are unicellular microorganisms often referred to as protists, which include the photosynthetic phytoplankton as well as some heterotrophic organisms (e.g. amoebae). Within this diverse collective of microorganisms, biodegradation of oil compounds is primarily associated with bacterial processes. In the marine environment the archaeal responses to oil pollution have mainly been considered low when compared to bacterial responses (e.g. Röling et al. 2004). However, archaea are important members of communities in petroleum reservoirs and are involved in the mineralization of petroleum hydrocarbons, for instance by methanogenesis under anoxic conditions. Recent studies have also implied the contribution of halophilic archaea to the biodegradation of hydrocarbons in the marine environment (Al-Mailem et al. 2010; Liu et al. 2009). Specific to the Arctic marine environment, essentially nothing is known of the archaeal contributions to oil degradation. 

Among eukaryotes, degradation of oil is mainly associated with fungi (Prince 2005). Only a few studies have been conducted to characterize fungal communities in Arctic marine environments (e.g. Gunde-Cimerman et al. 2003; Butinar et al. 2011). Among the microorganisms, our understanding of the protist contribution to oil biodegradation is also limited. Heterotrophic protists may affect petroleum biodegradation indirectly through their role as bacterial predators. Efficient grazing by protists can maintain the bacterial community in the log-growth phase, resulting in increased metabolic rates and thus increased biodegradation rates (Ballestero and Magdol 2011). 

In summary, most evidence for microbial degradation in the Arctic is connected to bacteria and bacterial communities. Our focus in this section will be on bacterial communities and the bacterial processes involved in oil biodegradation in the Arctic environment. However, the need for greater knowledge related to the contribution of archaeal and protistan organisms to oil biodegradation is an incentive to develop a better understanding of their contribution to oil biodegradation and microbial ecology in the Arctic. 

The Arctic Ocean represents a complex system of currents with influx and efflux of water (see Figure 5-1).  Cold and relatively less saline water enters the Arctic Ocean through the narrow Bering Strait between Alaska and Siberia, while warmer, more saline surface waters from the Atlantic penetrate the Arctic Ocean and are cooled as they move through the Greenland Sea and the Norwegian Sea. Water reaching the Arctic Ocean basin is swept into a huge circular current — driven by strong winds — the Beaufort Gyre. Siberian and Canadian rivers drain into the circular marine current to create a captured internal reservoir of relatively fresh water. Periodically, the circular current weakens, allowing large volumes of fresh water to leak out and cross the Arctic in the Transpolar Current. The water exits the Arctic Ocean via several “gateways.” It can flow through the Fram Strait, between northeast Greenland and Svalbard Island, and then branch around either side of Iceland. It can flow around the west side of Greenland through Baffin Bay and out Davis Strait. It can also flow through a maze of Canadian islands and out Hudson Strait. These mixes of salty and fresh water also generate Arctic haloclines, a vertical effect in which the cold fresh water lies atop warmer saltier water.

Figure 5-1. Prevailing currents in the Arctic Ocean (Source: Woods Hole Oceanographic Institution; http://www.divediscover.whoi.edu

The transport routes carry microorganisms into the Arctic from a complex mixture of seawater and river sources, with the North-Atlantic current (number 6 in Figure 5-1) as the single most important influx transport route. This results in cosmopolitanism of microorganisms in the Arctic Ocean and the presence of sometimes unexpected microorganisms in this cold marine environment (Hubert et al. 2009).

Significant to the discussion of currents, water masses, and effects on microbial population structures, a considerable part of the Arctic consists of one- or multi-year ice coverage. Since marine ice is frozen seawater the microbes frozen into the ice are transported with ice movement. This movement essentially follows two routes; the Beaufort Gyre circulating clockwise around the North Pole and the Transpolar Drift Stream, where ice moves from the Siberian coast of Russia across the Arctic basin, exiting into the North Atlantic off the east coast of Greenland (number 5 in Figure 1).  During ice growth, brine pockets are generated in the ice, since ice crystal generation is almost devoid of impurities (Petrich and Eicken 2010).  Brine accumulations are first generated as pockets, and later as chimney-like tubes termed brine channels. Microbes require a fluid environment for active metabolism, and these brine channels represent a fluid microenvironment within the ice mass. These channels therefore become important ecological niches for microbes which are able to survive in the high salinity and subzero temperatures in these systems. Eventually, contact with seawater will result in nutrient transport and reduced salinity in the channels. If channels are wide enough (approximately 2 mm in diameter), they are able to support counter-flow, i.e. simultaneous upward and downward movement inside the same channel (Lake and Lewis 1970).  Despite the potential for flow and mixing with bulk marine water, brine salinity tends to increase with distance to the seawater.   These salinity gradients in brine channels may result in gradients of microbial population density.  In winter, the apparent effect is higher bacterial and algal concentrations in the lower than the upper parts of the ice (Gradinger and Zhang 1997). However, in late spring and summer with rising air temperatures, the opposite situation is observed. The ice is warmed up from the top, becoming more porous, and melt water penetrates the porous ice and generates a linear bulk salinity profile that is commonly observed in the upper 20 to 50 cm of the ice during melt. The organisms in the brine channels are therefore exposed to fluctuating and physiologically challenging salinity conditions during the periods of ice freezing and melting (Faksness et al. 2011). Since much of the brines are drained off the ice during the first year after freezing, multi-year ice tends to have lower salinity and density than first-year ice (Tang et al. 2007), resulting in fewer brine pockets and also fewer possibilities for active microbial metabolism in the ice.

The population structures of bacteria in Arctic seawater are comparable to those in seawater from temperate regions, with the predominance of Alphaproteobacteria, Flavobacteria/Bacteroidetes,Gammaproteobacteria and Verrucomicrobia constituting more than 90% of the communities (Comeau et al. 2011; Teske et al. 2011; Ghiglione et al. 2012). It is also relevant to compare the microbial communities in the Arctic and Southern Oceans, since the two geographical areas have similar climatic conditions. Although some comparative studies of the microbial communities in Arctic and Antarctic marine environments indicated a high degree of resemblance (e.g. Bano et al. 2004; Brinkmeyer et al. 2003), a recent pole-to-pole study of surface and deep marine bacterial communities revealed significant differences, with 78% of operational taxonomic units (OTUs) unique to the Southern Ocean and 70% unique to the Arctic Ocean. Despite these dissimilarities, polar ocean bacterial communities were more similar to each other than to lower latitude pelagic communities (Ghiglione et al. 2012).

At a more localized scale, sea ice microbial communities may differ from those in the surrounding seawater. A study of first-year Arctic ice showed differences between ice and source seawater populations that were characterized by an increased abundance of Gammaproteobacteria and a lower abundance of Bacteroidetes in sea ice relative to seawater.  Winter sea ice communities were more similar to autumn water communities; species-specific die off processes were not observed in sea ice.  The implication was that winter sea ice communities are stable, whereas seawater communities undergo seasonal succession. The Alphaproteobacteria dominated in both environments (Collins et al. 2010). A study of multi-year ice revealed overall dominance of Gammaproteobacteria, as well as an increase ofBacteroidetes in the sea ice when compared to source seawater; the abundance of Alphaproteobacteria was greatly reduced in sea ice relative to the abundance in surrounding seawater (Bowman et al. 2012). A number of studies have also investigated the microbial communities in Arctic sediments, which may be regarded as pristine environments compared to most other sediments investigated. Studies from an Arctic fjord (Svalbard) showed that the Alphaproteobacteria predominated in the overlying seawater, while Gammaproteobacteria were more abundant in surface sediments. Deeper anoxic sediment layers were dominated by Deltaproteobacteria, which include common genera of strictly anaerobic bacteria associated with marine sediments (Ravenschlag et al. 2001; Teske et al. 2011).

The primary climatic condition associated with the Arctic is low temperature, and therefore most reviews and fundamental studies of Arctic microbiology focus on low temperature and psychrophilic (cold-loving) microbes. However, other conditions are also specific for this environment, including extreme light/dark conditions in the Arctic summer and winter, respectively. The Arctic area also includes coverage, with 9-12 million km² of pack ice with average ice thicknesses of 3-4 m, with ridges up to 20 m thick. As described above the microbes in this ice are not only challenged by the subzero temperature, but also by high salt concentrations in the brine channels.

The temperature of the Arctic Ocean normally does not exceed 4-5°C, with cold (close to 0°C) low-salinity surface water and warmer high-salinity water (up to 5°C) below 50 m depth. The marine organisms in the Arctic are therefore consistently exposed to low temperatures. Numerous studies of microbial adaptions to low temperature have been published (e.g.  Brakstad et al. 2008; Helmke and Weyland, 2004; Knoblauch et al. 1999; Robador et al. 2009). As the temperature decreases with concomitant decrease in the thermal energy of a system (enthalpy), increased stability and rigidity of biomolecules occur. Proteins become less flexible, while the secondary structures of DNA and RNA become more stable, and the processes of DNA replication, transcription of RNA and translation of proteins are inhibited (Bakermans 2012). However, psychrophilic microorganisms have adapted to these low-temperature conditions in a number of ways:

• Enzymes in bacteria isolated from sea ice may be activated by the cold, with detectable catalytic activity well below the freezing point of seawater (Groudieva et al. 2004). The enzymes of psychrophilic microbes are characterized by increased flexibility of active sites resulting from structural modifications to address freezing temperatures.  These modifications may include less enzyme core hydrophobicity, increased surface polarity and fewer interactions between amino acids (hydrogen bonds, salt and disulphide bridges). These characteristics facilitate retention of enzymatic activities at low temperature (Bakersman 2012).
• Nucleic acids show adaption to low temperature at the sequence level in psychrophiles, as exemplified by increased uracil content in ribosomal RNA and dihydrouridine content in transfer RNA as a result of decreased temperature (Khachane et al. 2005; Dalluge et al. 1997).
• Microbial membrane structures may crystallize at low temperatures, impairing function. However, in psychrophiles, membrane fluidity is retained by increasing the proportion of unsaturated fatty acids and by desaturation of membrane lipids, resulting in unsaturated fatty acyl chains and less tightly packed lipid chains (Russel 2008).
• Microorganisms can be challenged in ice by reduced water availability due to the formation of ice crystals. However, psychrophiles can combat the effects of reduced water activity by the synthesis of compatible solutes such as sugars, amino acids, alcohols or cryoprotective proteins. Ice crystal formation can also be prevented by production of anti-nucleating materials (Bakersman 2012).
• Psychrophilic bacteria in brine channels are challenged by both low temperature and high salinity.  These may produce extracellular polymeric substances (exopolysaccharides), which have a cryoprotective role in the sea ice brine channels, as well as binding essential cationic trace metals to enhance halotolerance (Nichols et al. 2005).
• Microbes exposed to large drops in temperature may produce cold shock and cold acclimation proteins (Berger et al. 1996). Cold shock proteins are supposed to be involved in protein translation regulation, while cold acclimation proteins show high catalytic activity at low temperatures and rapid inactivation at moderate temperatures (Fukunaga et al. 1999).

In Arctic regions, with a long polar summer and a correspondingly long polar winter and associated widespread ice coverage, phototrophic organisms must tackle a number of challenges.  For example, phytoplankton communities responded rapidly and increased in numbers by an order of magnitude in the Beaufort Sea with the increased solar irradiation in May, despite the presence of sea ice. However, eukaryotic phototrophs somehow also persisted throughout the winter darkness (Terrado et al. 2008). Among phototrophic bacteria isolated from the Arctic Ocean were photoheterotrophic microbes, which are capable of both utilizing dissolved organic materials and harvesting light energy.  Photoheterotrophic taxa included coccoid cyanobacteria (Synechococcus and Prochlorococcus), aerobic anoxygenic phototrophic (AAP) bacteria, and proteorhodopsin (PR)-containing bacteria. In a study of photoheterotrophic microbes in the Arctic Ocean in the summer and winter it was observed that some of these bacteria (AAP and PR containing bacteria) decreased from summer to winter, in parallel with a threefold decrease in the total prokaryotic community, while the abundance of Synechococcus did not decrease in winter (Cottrell and Kirchman 2009). These data demonstrated the ability of phototrophs to switch from photosynthesis to heterotrophic activities during the winter.  Characterization of green light-absorbing PR from bacterial isolates from the Arctic Ocean showed a slower photocycle, relatively larger change in activation energy of the transition between the oxygen intermediate and the ground state compared to PR from bacterial phototrophs isolated from temperate seawater (Jung et al. 2008).

Photosynthesizing microbes have also been observed within sea ice. Marine algae seem to be more sensitive to brine salinity than bacteria and therefore are detected mainly in the lower parts of sea ice (lower salinity in winter associated with contact with seawater; Gradinger and Zhang 1997). Recent studies have also identified marine autotrophic bacteria in marine ice indicating the potential ability for ongoing use of dissolved organic materials (Díez et al. 2012).

Conditions in marine ice vary considerably from those in seawater. The temperature is lower in the ice relative to fluid water, and may drop as low as the air temperature during the polar winter. Transmission of light through the ice is severely reduced compared to seawater, and light may be entirely absent when the ice is covered by a snow cap. In first-year ice, brine inclusion networks are generated by salting-out processes, as described in previous sections. The brine channels of the ice provide liquid saline niches at subzero temperatures which enable microbial motility and metabolic activity (Junge et al. 2002; 2003; 2004; 2006). The brine channels also act as a matrix for transport of soluble organic matter and other nutrients.  In the brine channels the ice temperatures can be as low as -20°C and salinity conditions can be as high as 200 Practical Salinity Units (Salinity in normal seawater is 35 PSU). Though challenging to microbial life, laboratory experiments have shown that bacteria (Colwellia psychroerythraea) were able to incorporate amino acids into proteins at temperatures as low as -20°C in brines (Junge et al. 2006). This metabolic activity was associated with particles or with the surfaces of the brine channels (Junge et al. 2003; 2004). In addition, the ability to grow at subzero temperatures has been demonstrated in the bacterium Psychromonas ingrahamii, which was cultured at temperatures as low as ‑12°C (Breezee et al. 2004). The brine channels may become depleted in oxygen, for instance if microbial processes occur without oxygen renewal (seawater or photosynthesis), and anaerobic denitrification processes have been measured in marine ice (Rysgaard and Glud 2004). These studies therefore demonstrate that bacterial activity can take place in marine ice, both with respect to motility within the ice and active metabolism. Recently, Mykytzuk et al. (2013) presented a detailed characterization of the physiological adaptations that enable a halotolerant psychrophilic permafrost bacterium to grow and metabolize at temperatures as low as -15°C.  The implication is that halotolerant psychrophilic species may also exist in the ice.

Since Arctic environments are increasingly exposed to petroleum exploration, production and transport, the studies of microbes and processes involved in cold-environment biodegradation is essential both for the industries operating in these areas and the governmental bodies responsible for environmental stewardship.  As described, microbes can function within most Arctic environments, including ice at sub-zero temperatures.

Numerous studies have shown that populations of known oil-degrading bacteria are present in Arctic environments (Zinger et al. 2011, Ghiglione et al. 2012; Sul et al. 2013).  However, few studies have directly assessed microbial community composition with the ability to clearly distinguish all of those taxa that can be defined as oil-degrading bacteria (e.g. heterotrophic and other non-designated microbes that respond positively to oil). 

The physico-chemical characteristics and weathering conditions of different oils at low temperatures and which are treated chemically may vary, having impacts on biodegradation efficiencies and should be an issue for further research, for instance, in the relationship between biodegradation and oil appearance (e.g. viscosity, dispersibility, resurfacing) after a spill.  A variety of processes in the Arctic are season-variable (e.g. photooxidation, different ice-conditions).

Marine ice poses a particular challenge to the indigenous microbes, which require the ability to survive and be metabolically active at sub-zero temperatures and at high salinity.  Microbial metabolism has been demonstrated in marine ice, and populations have been stimulated by oil in the ice.  However, the extent of hydrocarbon biodegradation in Arctic ice is likely to be low but requires more attention.

Few studies have attempted to use modeling approaches to predict oil biodegradation in cold marine environments.  Modeling tools are used today by oil companies and authorities to predict the fate of oil spills, but these models are often less well calibrated for the Arctic environment.    In that respect temperature-related biodegradation data, which often are based on Q₁₀-approaches may show erroneous results at very low temperatures, probably due to physical changes in the oil.  Therefore, more detailed temperature-related biodegradation studies will be required to improve fate models, which often rely on incomplete data sets for cold climate spills. 

Typical oil discharge scenarios include process losses or accidental releases from exploration, production, processing or transport related activities. However, the risk of accidental oil discharges will always be present in an environment with oil-related activities. In addition, the diminishing Arctic ice zones will likely increase ship activity through the Northwest and Northeast passages, activities which may contribute to an increased risk of oil discharges in Arctic marine environments. Biodegradation of oil compounds is regarded as the most complete process, as it is able to completely remove oil compounds, through the mineralization to carbon dioxide and water. In particular, two large oil spills in marine water have drawn public attention to oil biodegradation issues in general, the Exxon Valdez grounding in Prince William Sound in 1989 and the recent Deepwater Horizon blowout in 2010, both recently reviewed by Atlas and Hazen (2011). Media attention on the deep water plume from the Deepwater Horizonblowout has resulted in increased focus on the potential effects of oil spills in cold water environments. 

Physico-chemical properties of the spilled oil will affect biodegradation. Crude oils may be separated into 4 main types; paraffinic, asphalthenic, naphthenic and wax-rich (Moldestad et al. 2003). For instance, paraffinic oils will have a distinct chromatographic n-alkane pattern and a high content of light compounds like BTEX and naphthalenes. These oils therefore contain a high degree of easily biodegradable compounds. However, naphthenic oils are often biodegraded in-reservoir (Head et al. 2003), and in these oils most of the easily biodegradable compounds may have been consumed before oil production.  Asphalthenic oils are rich in poorly biodegradable asphaltenes and resins, while wax-rich oils will often have a high pour point and may be solidified in cold water, a physical factor that will affect their availability for biodegradation.  Marine fuel oils may also be spilled.

When oil is discharged to the marine environment a number of weathering processes occur:

Volatile compounds with low boiling points (e.g. saturates up to nC₁₁, mono- and some diaromatic hydrocarbons) are rapidly evaporated after surface spills. These are compounds that are normally rapidly biodegradable in the water column, but evaporation is normally more rapid than biodegradation after a surface spill. Evaporation is slower in cold than in temperate seawater (Brandvik et al. 2005), and this may result in temporarily higher concentrations of volatile toxic compounds (e.g. BTEX) in the seawater. At high concentrations these compounds may prolong microbial lag-phases and delay the onset of biodegradation (Atlas and Bartha 1972; Hokstad et al. 1999), although it is not known if this will have an impact on biodegradation under field conditions. In subsurface releases these volatiles are rapidly dissolved from dispersed oil and we suspect may be biodegraded rather than evaporated. Evaporation also results in increased viscosity of the residual oil (Faksness 2008), which will negatively affect the ability of oil to disperse, thereby slowing biodegradation.

Components dissolved from the oil phase are available for biodegrading microbes in the water column. In cold seawater the dissolution of oil compounds is decreased compared to temperate water (Faksness, 2008). The typical oil compounds in a water-soluble fraction (WSF) from fresh oils include phenols, naphthalenes and 2-3 ring PAHs. In addition, the WSF contains considerable amounts of highly polar compounds with nitrogen, sulphur, and oxygen atoms in their structures (so-called NSO compounds), often present as a chromatographic "hump", termed the "unresolved complex mixture" (UCM). In a study of WSF from an in-reservoir biodegraded oil (Troll) approximately 70 % of the WSF was separated by preparative high-pressure liquid chromatography, into a polar fraction (Melbye et al. 2009).  The non-polar compounds of the WSF are often considered to be rapidly biodegraded in the marine environment (Brakstad and Faksness 2000), and biodegradation of these compounds may result in a significant increase in the UCM  concentration relative to other crude oil components (e.g. Meredith et al. 2000), highlighting their persistence (Han et al. 2008).

Photooxidation is an important process in degrading and transforming crude oil compounds after release to the environment. The polar region exhibits vast seasonal differences in light conditions, and as a result photooxidation varies significantly between the polar summer and winter. UV-irradiation of crude oils has shown that aliphatic compounds are mainly resistant to photodegradation, while aromatic compounds appear particularly sensitive to this process (Maki et al. 2001). In contrast to biodegradation, increased size and alkyl substitution result in increased sensitivity of aromatic hydrocarbons to photochemical oxidation. As photooxidation leads to the inclusion of oxygen atoms in the structures of these compounds, photooxidized products appear mainly in the polar resin fraction of the oil (Maki et al. 2001; Garrett et al. 1998; Prince et al. 2003). Additionally, the average molecular weight of oil compounds is reduced and the oxygen content increased.  Studies have also shown that the photooxidized compounds subsequently exhibit increased susceptibility to biodegradation (Dutta and Harayama 2000; Maki et al. 2001; Ni'matuzahroh et al. 1999). Consistent with the physico-chemical properties of the photooxidized compounds, both the dissolved organic carbon concentration and acute toxicity of the water-soluble fraction of oil increased during the irradiation period (Maki et al. 2001). Thus, photooxidation results in a greater proportion of oxidized compounds that exhibit increased water-solubility and subsequently more significant impacts on toxicity and biodegradation. However, investigation of the relationship between photooxidation, biodegradation and toxicity would be of interest as part of the fate-determination of different oil compounds during the Arctic summer.

In shallow seawater and at higher levels of suspended sediments (e.g. after a storm), sediment particles may adhere to the oil and sink to the subtidal seabed sediments. Oil spills may also drift to shore and be mixed into the intertidal sediments. Oils mixed into the sediments will be subject to microbial processes in the sediment (Refer to Section 3). If seawater replenishment is poor, aerobic processes may consume most of the oxygen, resulting in anoxic conditions. Anaerobic biodegradation of hydrocarbons, by several alternative mechanisms, will occur in the absence of oxygen, as reviewed by Heider (2007).  Genes associated with anaerobic hydrocarbon degradation (e.g. benzyl- and alkylsuccinate synthase genes) have been detected in hydrocarbon-contaminated sediments (Callaghan et al. 2010).  

Water uptake into the spilled oil may cause the formation of viscous and often stable water-in-oil emulsions. Emulsions have been shown to be poorly biodegradable (Brakstad et al. 2011; Cook et al., 2011). Water taken up in the emulsions may contain oil-degrading microbes, but if water is trapped in the emulsions this will not promote biodegradation on the bulk oil as the emulsions may be depleted in essential nutrients. 

With sufficient energy from wave action the oil may break up into droplets in the water column. If oils are easily dispersed, small droplets are generated. The rising or settling rate of the droplets is related to the size and specific gravity of particles. As an example droplets 100 µm in diameter and a lower specific gravity that the surrounding seawater have been observed to rise with a velocity of approximately 1.5 m/h. Thus, larger droplets will resurface rapidly and thin oil films (sheens) may be formed. Oil dispersion is important for biodegradation. Dispersible oils will generate relatively small droplets, resulting in large surface areas for bacterial attachment. For instance, fresh Louisiana Sweet Crude oil can be made to generate dispersions with a median droplet size of 50-150 µm under continuous breaking wave conditions in an oil-on-seawater flume experiment (Brakstad et al. 2011). Several biodegradation studies of dispersed oil in cold seawater (5-8°C) have shown bacterial colonization of oil-droplets and biodegradation of dispersible oils (e.g. Lindstrom and Braddock 2002; MacNaughton et al. 2003; Venosa and Holder 2007; Prince et al. 2012). This colonization may tend to generate flocs of oil and biomass (MacNaughton et al. 2003; Bælum et al. 2012). However, for waxy oils with high pour points, evaporation, dilution and dispersion may be reduced in cold seawater, since precipitated wax may form a matrix which limits internal mixing and acts as a diffusion barrier between the oil and the water (Faksness 2008).

As described above, surface and resurfaced oil may generate thin films on the sea surface. In a series of studies with thin oil films immobilized on hydrophobic adsorbents, n-alkanes in these films were rapidly biodegraded in temperate and cold seawater (0-13°C), while aromatic compounds were subject to mixed dissolution and biodegradation (e.g. Brakstad and Bonaunet 2006; Brakstad et al. 2004). In experiments over a period of 112 days with different oil thicknesses of a wax-rich oil it was apparent that a thickness limit for measurable biodegradation (nC₁₇/Pristane and nC₁₈/Phytane) was between 0.1 and 1.0 mm in cold (6-10°C) seawater (Brandvik et al. 2006). 

In the aftermath of the Deepwater Horizon incident, a large body of new information has been collected and integrated with our already existing understanding of the microbial response to oil spilled in the marine environment (Hazen et al. 2010; Mason et al. 2012; Valentine et al. 2012). In general, in situ sampling and analysis revealed unexpectedly rapid disappearance of released oil in the Gulf of Mexico environment, which is characterized by a temperate climate (Hazen et al. 2010). This rapid disappearance was affected by the prevalence of water-soluble constituents in the crude oil (Reddy et al. 2012), injection of subsea dispersant into the erupting oil flow (Kujawinski et al. 2011), and presence of indigenous oil-degrading microorganisms in this area that is well known for natural seeps of crude oil from reservoirs (Lu et al. 2012). Such indigenous oil-degrading microorganisms are the topic of this section.

Following the Deepwater Horizon incident, extensive analysis of microbial responses was done both in situ and in laboratory microcosms. These analyses support, in general, a paradigm of successive blooms of taxonomically distinct indigenous microbial populations as the oil weathers and labile components are sequentially degraded leaving less-readily degraded components to feed subsequent blooms (Hazen et al. 2010; Valentine et al. 2010; Kostka et al. 2011; Baelum et al. 2012; Beazley et al. 2012; Lu et al. 2012; Mason et al. 2012; Valentine et al. 2012).

Conditions are very different in high latitude marine environments.  As described in previous sections, the Arctic and Antarctic marine environments are characterized by seasonal extremes of photoperiod, spatial variability in salinity and temperature, as well as generally colder surface temperatures compared to the temperate latitudes. These differences may result in different expectations about the rate of oil degradation, as described in previous sections.  They also result in different expectations about the indigenous populations of oil degrading microorganisms.

As mentioned in previous sections, microbial responses to oil in marine environments generally are dominated by bacteria rather than archaea (Roling et al. 2004). Although fungi are known to degrade petroleum compounds in some marine settings (Zinjarde and Pant 2002), few surveys of fungal abundance in high latitude marine environments have been done (Butinar et al. 2011) and thus far none have addressed oil degradation by fungi in high latitude environments. For these reasons, this section focuses on the bacterial component of the marine microbiological community. 

Among the bacterial taxa catalogued in high latitude marine environments, many appear to be specific to that environment (Ghiglione et al. 2012; Sul et al. 2013).  This apparent specificity may be due to truly unique populations, or it may be a function of the limit of detection.  Community members that thrive in the high latitude marine environment grow to relatively high cell densities and are therefore more easily detected.  Various investigations have found that microbial species richness curves are not saturated with typical levels of effort.  This finding has led to the hypothesis that there is an under sampled “rare biosphere” of organisms with low population density (Sogin et al. 2006) that, despite low population levels, can respond to changes in environment and energy source.  This phenomenon may be typified by the explosion of Oceanospiralles and Colwellia populations in the presence of different partitions of spilled oil during the Deepwater Horizon incident (Hazen et al. 2010; Bælum et al. 2012).

Marine ice represents an extreme biosphere with below-zero-centigrade temperatures and high salt concentrations. It has been demonstrated by field studies that bacterial populations in Arctic marine ice are affected by oil pollution, stimulating species of a few genera like Colwellia, Marinomonas and Glaciecola (Brakstad et al. 2008). 

Not all of the microorganisms found in the Arctic oceans are adapted to that environment.  The various currents carry viable microorganisms from diverse locations to the Arctic (Rosnes et al. 1991; Hubert et al. 2009; Hubert et al. 2010); thus, there is an expectation of cosmopolitanism among the free-living microorganisms.  This is not to say that the population structure is homogenous as if the Arctic were a giant mixing bowl.  In fact, there is documented variability in population structures, with different communities associated with water masses of different origins (Galand et al. 2010; Sul et al. 2013).   The presence of non-adapted microorganisms such as thermophiles does, however, indicate that microbial populations adapted to the consumption of natural or human-induced oil releases might be transported to and be present in areas that are not commonly exposed to oil.

Crude mineral oil is degradable by indigenous microorganism populations in the Arctic marine environment, even at near-freezing temperatures (Brakstad and Bonaunet 2006), although at slower rates compared to higher temperatures (Margesin et al. 2003; Michaud et al. 2004).  Nevertheless, over a time course on the order of weeks substantial biodegradation can be observed in nutrient-enriched cold Arctic seawater (Brakstad and Bonaunet 2006). Community analysis of oil-degrading Arctic microbial consortia indicated that several taxa of bacteria are involved in biodegradation in this environment, including genera related to Pseudoalteromonas, Pseudomonas, Shewanella, Marinobacter, Psychrobacter, and Agreia (Deppe et al. 2005).  Of interest, these are different organisms from those directly associated with degradation in the Deepwater Horizon spill in the Gulf of Mexico, specifically bacteria of the orders Oceanospiralles (Hazen et al. 2010; Kostka et al. 2011) and Alteromonadales (Bælum et al. 2012), among others.

Linear alkanes often are characterized as an easily accessible carbon source, either through degradation or direct incorporation into microbial biomass, in the marine environment (Harayama et al. 1999).  The metabolic pathways for linear, branched, and cyclic alkanes have been studied and described since the 1960s (Jobson et al. 1972, Coates et al. 1997, Feng et al. 2007, Rojo 2009, Gray et al. 2011).  Preferential degradation of short-chain alkanes (represented by C₁₅) over long-chain alkanes (represented by C₂₆) was observed in situ in a deep plume (circa 1,400 m) in the Gulf of Mexico under aerobic conditions (Hazen et al. 2010).  Furthermore, during weathering in subsurface petroleum reservoirs, alkyl chains on substituted soluble PAHs such as alkane-substituted naphthalenes may be transformed even more rapidly than linear alkanes (Jones et al. 2008).  Whether this phenomenon, observed in anaerobic subsurface reservoirs, would occur in the presence of petroleum hydrocarbons released into the deep sea, remains unknown. 

The specific bacteria known to accomplish alkane degradation are numerous (Whyte et al. 1997; Rabus et al. 1999, Hara et al. 2003, van Beilen et al. 2004, Throne-Holst et al. 2006, Feng et al. 2007, Throne-Holst et al. 2007, Wentzel et al. 2007, Rojo 2009, Teramoto et al. 2009, Wasmund et al. 2009, Tapilatu et al. 2010, Alonso-Gutierrez et al. 2011, Teramoto et al. 2011).  Among these, many were characterized from high-latitude marine environments.  Specifically, the Pseudomonas strains isolated by Whyte et al. (1997) from Arctic soils may be transported to the marine environment via runoff.  Alcanivorax species are known to be widespread in marine environments exposed to oil (Hara et al. 2003, van Beilen et al. 2004) and, if not prevalent in the Arctic environment, might be expected to be present because of currents.  Thus, bacteria capable of alkane degradation are expected to be present in the Arctic oceans.

The ability to degrade aromatic hydrocarbons and, in particular, polynuclear aromatic hydrocarbons typically is considered to be less widespread than the ability to degrade alkanes.  For example, some organisms have diverse pathways that confer the ability to degrade polynuclear aromatic hydrocarbons, e.g., Mycobacterium vanbaalenii (Kweon et al. 2011) and various Pseudomonas spp. (Whyte et al.1997). The distribution of these genes among bacteria in Arctic marine environments remains unknown.       

In general, biodegradation of oil compounds is expected to follow the order n-alkanes > branched alkanes > low molecular weight aromatics > cyclic alkanes (Perry 1984). In cold seawater the same order is expected, although degradation will be highly influenced by the physico-chemical characteristics of the oil. The low temperature affects both dissolution from the non-aqueous (crude oil) to the aqueous phase (Schluep et al. 2001), and evaporation of volatile compounds, as described above.  

At temperatures above the freezing point of seawater (approximately -1.8°C) biodegradation of crude oil hydrocarbons is well documented. This is exemplified in Figure 5-2, showing the mineralization curves of ¹⁴C-labelled naphthalene, phenanthrene and hexadecane in seawater at 0°C when the compounds were spiked into crude paraffinic oil. Degradation of the n-alkane (hexadecane) was faster than for the aromatic compounds, and a smaller aromatic (naphthalene; 2-ring) degraded faster than larger aromatics (phenanthrene; 3-ring). This pattern followed the generally accepted order of crude oil compound biodegradation described above.

Figure 5-2. Mineralization in seawater at 0°C of 14C-labelled hydrocarbons spiked into crude oil (from Brakstad and Bonaunet 2006). No mineralization was measured in sterile controls (not shown).

One of the first attempts to study oil biodegradation in Arctic seawater at low temperatures (2-11°C) showed that shifts in microbial populations towards more oil-degrading bacteria, that abiotic oil losses were lower than expected, and that various classes of hydrocarbons (saturates, mono-, di- and polyaromatics) were subject to biodegradation (Horowitz and Atlas 1977). Several studies have compared oil biodegradation in seawater or with bacterial cultures at different temperatures, and results from some of these including temperatures relevant for the Arctic are summarized in Table 5-1.

In summary, these and most other relevant studies (e.g. MacNaughton et al. 2003) show slower biodegradation by lowering of the temperature, but the results also show that biodegradation at low seawater temperature is considerable. In a recent study with low concentrations (2.5 mg/L) of Alaska North Slope oil with Atlantic seawater, 80% was biodegraded (saturates, 2- to 4-ring aromatics) after 60 days at 8°C (Prince et al. 2012). While laboratory studies indicate that biodegradation in Arctic seawater may be slower than in temperate seawater, these results have not been confirmed by field studies. Seasonal biodegradation data and comparison of oil biodegradation from different geographic areas with the same oils and analytical procedures may be necessary to test these assumptions. Oil characteristics should also be addressed in more detail, for instance, by comparison of dispersed oil biodegradation of different oil types and weathering degrees at several seawater temperatures.  The physical properties of oil may decrease bioavailability of oil (e.g. larger droplets at lower temperatures would increase the surface area-to-volume ratio).

Table 5-1. Summary of selected biodegradation studies performed at different seawater temperatures 

^A) k1 is first-order rate coefficient; t_½ is half-life (0.69/k1) ^B) Reduction determined by comparison to sterile controls

Several biodegradation studies of oil in Arctic sediments have been conducted, most of these to investigate the potential for bioremediation of stranded oil in the Arctic (see later chapter). Studies on oil pollution of Arctic and Antarctic beaches has demonstrated the presence of indigenous hydrocarbon-degrading bacteria in these pristine environments (e.g. Grossman et al. 2000; Delille and Delille 2000; Powell et al. 2005). Oil removal from beach sediments may be attributed to several processes, including physical removal, photooxidation and biodegradation. For instance, significant depletion of total hydrocarbon concentrations and mineralization of radiolabelled hexadecane have been measured in Canadian Arctic soils at 4°C (Greer 2008). Anaerobic biodegradation has also been measured in the Arctic. Low-temperature degradation of PAH-compounds was reported from Arctic soils under anoxic and nitrate-reducing conditions at 7°C (Eriksson et al. 2003). In the Arctic winter the upper parts of marine sediments become frozen.  Whether biodegradation stops or continues at very slow rates under these conditions is not known, although microbial activity at subzero temperatures has been demonstrated (Doyle et al. 2012).  Several studies with oil-contaminated freeze-thaw cycled soil or permafrost have shown that microbial respiration takes place even at subzero temperatures and hydrocarbon degradation was observed (Rike et al. 2003; Børresen et al. 2007; Chang et al. 2011).  These studies therefore demonstrate that the lower limit for biodegradation can be below the freezing point.

If oil spills reach the marginal ice zone, the ice may become oil-infested. Once trapped within the ice, ocean currents can transport the oil over large distances.  A secondary discharge situation occurs during the spring melt season and, if the ice has been transported from the original spill site, this can result in contamination of new locations. In the spring and summer seasons, chemical alteration of the crude oil through photooxidation may also become an important process (Refer to Section 3).  Although the immediate impact of oil spills in ice has been studied (e.g. Fingas and Hollebone 2003) and is fairly well understood, little is known about the long-term fate and effects of such pollutants on ecosystems in polar environments.  To date, few studies have attempted to determine the transport and fate of individual water-soluble oil components in sea ice.  However, data from some recent studies have shown that the more water-soluble compounds (mainly naphthalenes, phenanthrenes and dibenzothiophenes) migrate through the brine channels in the ice (Figure 5-3).  As a result, such compounds come into contact with sea ice microbes in the brine and the underlying water (Faksness and Brandvik 2008a; Faksness and Brandvik 2008b).

In line with the results from studies with Arctic soils one should expect that biodegradation may also take place in marine ice at subzero temperatures. As described earlier, microbial metabolism and motility have been measured in the brine channels of marine ice (Breezee et al. 2004; June et al. 2002; Junge et al. 2003; Junge et al. 2004; Junge et al. 2006; Mykytzuk et al. 2013).  However, biodegradation of oil in marine ice has not yet been fully investigated. In a winter field study (February to June) performed on Svalbard with crude oil frozen into fjord ice, a slow reduction in the ratio between naphthalene and phenanthrene was measured in the parts of the ice with downward migration of soluble compounds, while no significant change in n-C₁₇/Pristane was measured, as shown in Figure 5-3 (Brakstad et al. 2008). However, the bulk oil stimulated bacterial biomass, including a few bacterial genera expected to be oil-degraders (Brakstad et al. 2008). The results from another field study performed at Svalbard showed that no significant degradation of oil hydrocarbons occurred in the ice at subzero temperatures, but at 0°C melt pool oil samples fertilized with inorganic nutrients showed a significant change in bacterial diversity (Gerdes and Dieckmann 2006). Marine ice represents an extreme environment for life.  The combination of low temperature and high salt content in the brine channels require that microbes be both halo- and psychro-tolerant.  Extremely halophilic or halotolerant microbes able to degrade oil have been reported (e.g. Diaz et al. 2002, Al-Mailem et al. 2010), but not so far in cold environments.  However, as described above, it has been demonstrated that oil pollution in marine ice may stimulate the growth of a few specific bacteria (Brakstad et al. 2008), but the ability to degrade oil compounds needs to be clarified. In addition, most oils will also be solidified under these conditions, but the migrating water-soluble compounds may be relevant target compounds for oil-degrading bacteria in this environment. If this is true, bacteria able to degrade small aromatics may be more relevant than alkane-degrading bacteria. 

Figure 5-3. Oil migration and degradation in marine ice. [The middle core shows the migration of different oil components through an ice core with the chromatograms of the components in the left panel and the relative ratios of specific components in the right panel (from Brakstad et al. 2008)].

Several oil spill models have been developed during the last decades.  Most of these are physical models which can be separated into oil weathering models, trajectory models (predicts the route of an oil spill), or stochastic models (describing an impact area of an oil spill).  Examples of well-known models are the Oil Spill Contingency and Response (OSCAR) and the OILMAP models (www.sintef.no/Materialer-og-kjemi/Marin-miljoteknologi/Miljomodellering/Modellverktoy/OSCAR-Oil-Spill-Contingency-And-Response/; http://www.asascience.com/software/oilmap/index.shtml). Models have also been presented for predictions of oil behavior in ice-infested water (Drozdowski et al. 2011). Most of these models are physical models, but the OSCAR model also incorporates biodegradation of 25 pseudo oil compound groups, in addition to descriptions of the physical environment, physical-chemical fate processes and ecotoxicity (Aamo et al. 1997; Reed et al. 2000). In the OSCAR model, which is an industry standard in Norway, biodegradation is one of the fate processes together with physico-chemical processes like advection, spreading, evaporation, dispersion, dissolution, particle adsorption/dissolution, volatilization from water column, and seabed contamination. An example of vertical oil concentrations and mass balance after a simulated 60-day blowout is shown in Figure 5-4. However, biodegradation as part of the mass balance may be overestimated in the model, since degradation is determined on the bases of biotransformation, not complete biodegradation, and only compounds determined by gas chromatography-mass spectrometry (GC-MS) analyses are included.  

Figure 5-4. Simulation of a deep water blowout by the OSCAR model. [The left figure shows the oil concentration vertically in the water masses during a simulated 60-d blowout from 1600 m depth with a light paraffinic oil. The right figure shows the mass balance between different fate processes during the blowout period (from Brakstad et al. 2011).]

In the oil spill models biodegradation must be predicted at different temperatures. Oil biodegradation data in Arctic environments with cold seawater are limited, since most published studies have been performed at higher temperatures than relevant for these environments. In order to transform degradation data between different temperatures, plots have been used to transpose results of bacterial metabolism and growth at different temperatures. Temperature-related bacterial growth rates may be estimated by using modified Arrhenius plots (Arrhenius, 1889).  Ideally, Arrhenius plots should show temperature-related linearity.  However, in a study with psychrotrophic toluene-degrading strains of Pseudomonas putida grown on toluene or benzoate, growth rates had to be fitted using two linear segments at a temperature range of 4-30°C: one segment above and one below 17-20°C (Chablain et al. 1997).  When using Arrhenius plots, the temperature range should therefore not be too broad.  For water-soluble compounds, temperature-dependent biodegradation has been suggested to follow a Q₁₀-value, which is a relationship describing the degradation rate increases when temperatures are raised by 10°C increments.

Equation 1

Where R is the general gas constant (8,314·10^-3 kJ/mol·K), E_a is the activation energy (kJ/mol), T₁ is the reference temperature in Kelvin and T₂ is the actual temperature in Kelvin.  According to this approach, the degradation rates should double for every 10°C increase, resulting in an ideal Q₁₀ of 2.0.  The Q₁₀–values for the biodegradation of oil hydrocarbons in seawater were determined with a heavy fuel oil (Bunker C), and with winter or summer water samples from the North Sea.  When incubation temperatures of 4-18°C were used, Q₁₀–values of 2.4 and 2.1 were determined for waters in winter and summer, respectively, where biodegradation was measured as biological oxygen demand (Minas and Gunkel 1995).  Calculations of Q₁₀-values from a variety of studies have shown that the rule-of-thumb value (Q₁₀ = 2.0) is a fairly good approximation in a temperature range of 5–27°C (Andrea Bagi, personal communication).  However, for the narrow range and freezing temperatures of the Arctic the expectation that calculated Q₁₀ would remain close to 2 may not be valid.  For instance, a calculation of Q₁₀ in immobilized oil films (Statfjord B oil) based on data for 5 and 0°C (Brakstad and Bonaunet 2006) showed a value of 16.2 (Andrea Bagi, personal communication).  This may be caused by changes in the physico-chemical characteristics of this oil at these temperatures.  Thus, changes in oil characteristics at low seawater temperatures may affect the biodegradation models, and therefore predictions of oil degradation rates in Arctic seawater will require closer examination. 

Since oil consists of thousands of different compounds (Marshall and Rogers 2003) measurements of individual compounds is a challenge.  Bulk oil biodegradation may be determined by traditional gravimetric analyses (e.g. Horowitz and Atlas 1977), while broader groups of oil components (saturates, aromatics, resins and asphaltenes = SARA) may be determined by Iatroscan thin-layer chromatography with flame ionization detection (TLC-FID; Stevens 2004).  Using this method, crude oil components are determined according to their polarity.  The saturate fraction consists of nonpolar material including linear, branched, and cyclic saturated hydrocarbons (paraffins). Aromatics, which contain one or more aromatic rings, are slightly more polarizable.  The remaining two fractions, resins and asphaltenes, have polar substituents.  Additional bulk oil analytical methods include Fourier Transform Infrared (FTIR) spectroscopy and Nuclear Magnetic Resonance (NMR) spectroscopy.  FTIR is an absorption technique that uses infrared (IR) electromagnetic radiation to examine the identity of chemical bonds within the substance of interest.  As microbial degradation of the oil is expected to result in the addition of oxygen atoms into the structure of oil compounds this method may be a method for measuring bulk changes in composition, although the resolution and sensitivity is poor compared to other methods.  NMR is a nondestructive technique that is well-suited for identifying and quantifying different hydrocarbon classes and can provide information on the relative content of aliphatic, olefinic, and aromatic components.  Studies have shown that NMR spectra in conjunction with multivariate statistical analysis can be correlated to a number of physicochemical properties and standard distillation cut yields (Molina et al. 2007).  Mass spectrometry (MS) has become one of the most important detection principles in modern analytical chemistry.  The principle behind MS is that molecules can be identified through their molecular weight and fragmentation patterns.  MS is very often connected to a separation step, usually gas (GC) or liquid (LC) chromatography.  These methods may be used to identify and quantify targeted oil compounds or for fingerprinting of complex chemical mixtures.  To separate between different oil compound groups gas chromatographic analyses (GC-FID and GC-MS) are the standards today, but these methods favor detection of nonpolar compounds.  The common use of these methods therefore limits our knowledge of oil biodegradation, mainly to some compound groups, such as the C10-C40 saturates, cyclic saturates (decalines), BTEX, phenols, 2-6 ring PAHs, and a variety of biomarkers.  LC-MS analyses may therefore be an important supplement to the gas chromatographic analyses for more polar compound groups.  In addition, biodegradation studies of compounds like naphthenic acids have been of interest in specific areas like Canada.  Several high-resolution instruments, like time-of-flight mass spectrometers (ToF-MS) coupled to GCxGC systems (GCxGC-ToF-MS)(e.g. Tran et al. 2010) and Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer (e.g. Hughey et al. 2008) provide powerful techniques for the analytical separation of complex mixtures combined with methods for characterizing the resolved compounds. Minor components hidden in the large background can be detected by these instruments, and both resolution and sensitivity allow for searching of spectra from very narrow peaks.  For instance FT-ICR MS can separate masses of <0.002 Dalton of compounds that contain heteroatoms such as N, O, S and other elements, identifying oil compounds by mass and molecular formula at high resolution.

Advances in microbial sampling capabilities, in particular sampling of the ocean in drilling areas, came with advances in drilling technology.  The Ocean Drilling Project (ODP) and subsequent Integrated Deep Ocean Drilling Program (IODP 2003-2013) and planned International Ocean Discovery Program (IODP 2013-2023) provide a framework for these activities (Edwards et al. 2012).  Each of the named programs includes or will include a sampling component for microbial ecology research.  In particular, the 2003 IODP included extensive evaluations of seafloor and sub-seafloor microbial communities (Cyranoski 2003).  In conjunction with new microbiological techniques, these samples provided new perspectives on deep ocean microbial community composition and function (D'Hondt et al. 2004; Schippers et al. 2005; Inagaki et al. 2006; Biddle et al. 2008; Kobayashi et al. 2008; Forschner et al. 2009; Lomstein et al. 2012).  Understanding native populations in Arctic drilling fields requires sampling such as has been carried out in these programs.

Much of the sampling that is associated with drilling activities focuses on the deep subsea floor while water column and sediment samples can be collected with remote samplers or can also  be collected by autonomous underwater vehicles (AUVs).  The Chemosynthetic Ecosystem Science (ChEss) project of the Census of Marine Life (2002-2010) was one such project that generated a substantial amount of new information about marine microbial communities.  Much of the success of the ChEss project was attributed to the development of improved deep-ocean AUVs (German et al. 2011) that allowed systematic exploration of previously understudied areas, including cold seeps.  Modern AUVs are capable of rapid deployment and operation at a range of depths.  They have been effectively deployed to sample in response to events such as the Deepwater Horizon spill of 2010 (Camilli et al. 2010).  These vehicles contribute to the ability to observe natural processes and conduct in situ experiments, particularly at depth in harsh marine environments.

Another strategy is to employ microbial observatories in marine environments that incorporate real-time sensors, time-lapse cameras, and other experimental devices.  These observatories, along with autonomous and cabled sensors, allow direct measurement of microbial processes in the deep ocean.  In particular, beginning some 20 years ago, circulation obviation retrofit kits (CORKs) came into use to study connectivity of hydraulics and biogeochemistry at the interface of the ocean bottom and open water (www.corkobservatories.org; Cowen et al. 2003) rather than relying on extrapolation from controlled laboratory experiments (e.g. Bartlett 2002; Tapilatu et al. 2010) or inference from population composition (Simonato et al. 2006) as is more commonly done.  Another type of observatory, the Microbial Methane Observatory for Seafloor Analysis (MIMOSA), is an autosampler that collects and archives microbial material for later recovery and analysis.  Two of these devices recently were deployed in the Gulf of Mexico to evaluate petroleum seeps and spills as they affect microbial population structure (Balinski 2012).  This type of observatory may be useful to implement in situ experiments to monitor biodegradation rates and processes and further advance knowledge of petroleum hydrocarbon degradation processes in the deep ocean environment.

Finally, another tool that is directly relevant to petroleum biodegradation and carbon utilization is the so-called “bug trap,” in which hydrophobic beads or woven matrix is dosed with petroleum hydrocarbons to evaluate in situ degradation potential and analyzed to characterize degrading community composition.  Because petroleum-degrading microorganisms can be chemotaxic to suitable substrates, these experimental devices can be used to attract and study degraders in the laboratory (Brakstad and Bonaunet 2006) and in situ experiments (Raloff 2010; DeAngelis et al. 2011).

In standard laboratory studies, oil degradation is usually determined by comparison of depletion in normal seawater or cultures to depletion in sterile (killed) controls. In this way processes like evaporation, wall effects, dissolution of compounds from the oil phase etc. may be accounted for and separated from the biodegradation process. However, in field and meso-/large-scale studies biodegradation is determined by normalization of degradable compounds to less degradable (recalcitrant) compounds. Common compounds for this internal normalization are pentacyclic triterpane biomarkers (e.g.C₃₀17α(H),21β(H)-hopane) and the isoprenoids pristane and phytane (Prince et al. 1994; Douglas et al. 1996; Page et al. 1996). The isoprenoids have proven to be biodegradable themselves, although at slower rates than their corresponding n-alkanes (e.g. Douglas et al. 1996). Hopanes also have limitations if used to determine biodegradation of compounds with low boiling points, since it may be difficult to separate biodegradation from evaporation. In addition, determination of biodegradation as ratios between biodegradable and more persistent compounds has also been suggested using other compounds, like 2-methylphenanthrene/1-methylphenanthrene, and C3-phenanthrene/C3-dibenzothiophene (Fedorak and Westlake 1981; Christensen and Larsen 1993; Wang et al. 1998; Lamberts et al. 2008).

Most environmental studies of petroleum-derived chemicals have focused on effects related to specific and easily identified hydrocarbons such as n-alkanes, BTEX and PAHs. However, in the case of environmentally weathered samples, most oil compounds appear as an unresolved complex mixture (UCM in gas chromatograms, and are often referred to as the “hump” (see Figure 5-5). Having undergone a variety of weathering processes (e.g. evaporation, biodegradation and photooxidation), this residual UCM is comprised of thousands of environmentally persistent compounds (Gough and Rowland 1990, Killops and Al-Jaboori 1990). In fact, it has been established that natural biodegradation of spilled crude oil leads to a significant increase in the UCM  concentration relative to other crude oil components (e.g. Meredith et al. 2000), highlighting the persistence of these compounds. Fractionation and subsequent characterization studies have shown that both non-polar (e.g. aliphatic and aromatic) and polar (resin and asphaltene) compounds contribute to crude oil UCMs.  Both the aromatic hydrocarbon and polar UCMs have been shown to bioaccumulate in marine organisms and elicit ecotoxicological responses and impaired health (e.g. Farrington et al. 1982; Widdows et al. 1995; Barron et al. 1999; Smith et al. 2001; Rowland et al. 2001; Donkin et al. 2003). The polar UCM fractions comprise compounds containing highly polar N, S, and O atoms in their structures (so-called NSO compounds).  Many of these compounds (e.g. phenols and naphthenic acids) are thought to be homologous in structure to compounds present in the non-polar fraction of the UCM, hence their resistance to biodegradation.  Due to their persistence these compounds may reach other locations. 

Figure 5-5. Gas chromatogram of the aromatic hydrocarbon UCM fraction extracted from mussels (Mytilus edulis) from a petroleum impacted site (Southend) on the UK coast (From Booth et al., 2007a).

The biodegradation potential of these seemingly persistent UCM compounds is further complicated by the environmental conditions prevalent in Arctic regions.  Lower ambient temperatures will result in reduced biodegradation rates, and currently nothing is known about the abilities of psychrophilic or psychrotrophic bacteria to degrade these compounds. In temperate regions, microbial communities from previously oil-impacted sites have been shown to partially degrade model UCM compounds, such as alkylcyclohexyltetralins, alkylcyclohexylnaphthalene and naphthenic acids (Scott et al. 2005; Booth et al. 2007b; Frenzel 2008).  Recent studies using naphthenic acids showed that it was the molecular structure rather than the number of carbon atoms that was important for determining biodegradation.  Specifically, the most recalcitrant compounds included those with relatively high degrees of alkyl branching (Han et al. 2008).  During future degradation studies of oil compounds in cold environments it is therefore of major importance to consider these environmentally persistent and toxic UCM-related compounds.  

Naturally occurring hydrocarbon-degrading bacteria are found in all environments.  Although these bacteria are capable of initiating the biodegradation of spilled crude oil, attempts have been made to increase removal efficiency through bioremediation strategies.  Most bioremediation attempts have focused on developing good biostimulation strategies, typically by applying degradation rate-limiting nutrients, or the combination of these and other treatments, to accelerate the natural biodegradation processes.  Bioremediation processes, if successful, are cost-effective and reduce the environmental impacts of marine oil spills (Prince 1993; Swannell et al. 1996; Prince and Clark 2004; Prince 2005; Prince and Atlas 2005).  An alternative to biostimulation is bioaugmentation which involves the inoculation of indigenous or exogenous microbial cultures with high biodegradation potentials for contaminants.  Bioaugmentation approaches have been reported to improve biodegradation of hydrocarbons from oil spills in cold soil or marine sediments, and can be used in combination with fertilizers (Margesin and Schinner 1997; Ruberto et al. 2003).  Bacterial mats from marine oil-contaminated sites have also been suggested for use in the degradation of coastal oil spills, although these are of greater relevance for spills in temperate areas (Cohen 2002).  Various methods and strategies for bioremediation have been reviewed (e.g. Lee and Merlin 1999; Prince 2010).

To date, most experimental oil bioremediation studies in Arctic or Antarctic environments have been conducted on stranded oils, employing the application of fertilizers to stimulate the indigenous flora, and often in combination with mechanical treatments which improve oxygen and nutrient availability (e.g. Sveum and Ladousse 1989; Prince et al. 2003a; Obbard et al. 2004; Pelletier et al. 2004). Bioremediation strategies have also been applied to real oil spill situations such as the Exxon Valdez accident in 1989, where it formed part of a beach cleaning strategy (Bragg et al. 1994).

Also the use of chemical dispersants is regarded as an effective way of stimulating biodegradation of oil.  The dispersants work as surfactants, changing the surface characteristics of the oil, and reducing the droplet size.  Chemical dispersants have been used primarily on surface oil spills, but were also injected directly (Corexit 9500) at the wellhead during the Deep Water Horizon (DWH) spill (Atlas and Hazen 2011).

Biostimulation includes the addition of nutrients or other methods to enhance the capability of the indigenous microbial communities to degrade pollution components.  Biostimulation has been regarded as a cost-effective strategy for secondary cleanup mainly of stranded oil pollution.  However, other methods may also be regarded as biostimulatory actions, for instance the use of chemical dispersants for oil spills in seawater, since this may aid in oil biodegradation by increasing the oil surface area accessible to the oil-degrading microbes.

Most biostimulation activities have focused on stranded oil, with application of fertilizers to increase natural degradation by the indigenous microbial flora.  Biostimulation treatment is often combined with mechanical treatment to improve oxygen and nutrient availability.

In marine environments some growth- and biomass-stimulating factors are essential for oil biodegradation, especially nitrogen and phosphorous and the addition of these nutrients are common practice in bioremediation.  Balanced nutrient availability is important for biodegradation and the composition of hydrocarbon-degrading communities, since nutrient amendments, in some instances, can inhibit microbial activity (Braddock et al. 1997).  It is therefore important to avoid excess nutrients, which can cause detrimental effects, such as eutrophication.  During biostimulation, molar carbon/nitrogen/phosphorous ratios of 100/10/1 have often been used (e.g. Bouchez et al. 1995; Obbard et al. 2004).  However, results from laboratory studies have also shown that certain microbial populations may require different N/P ratios for optimal degradation of different hydrocarbons (Smith et al. 1998). Nutrient products are available as briquettes, granules or liquid fertilizers.  Liquid inorganic fertilizers have proven effective but require frequent application, and therefore oleophilic slow-release nutrient formulations have been developed, which promote hydrocarbon degraders at the oil-water interface.

For improved results bioremediation may be combined with other clean-up procedures.  Surf washing and the use of surfactants may increase the surface area of the oil and hence increase oil degradation. Ex situ technologies like land farming (spreading the polluted sediments over a larger area for better oxygenation), composting and biopiling may be used for treating oily waste during spill treatment (Lynch and Moffat 2005) although these approaches have seen limited application in polar environments.

Several field biostimulation trials have been conducted on Arctic beaches at Spitzbergen, either in Ny Ålesund (78° 55' N, 11° 56' E) or in the Van Mijen fjord close to the small mining society Svea (77°56’N, 16°43’E).  Experiments performed by SINTEF and the oil company ELF in the 1980s in Ny Ålesund with the slow-release oleophilic fertilizer Inipol EA22 indicated that application of the fertilizer to oil in beach sediments resulted in increased biodegradation in coarse sediments, but not on oil in finer sediments (Sveum and Ladousse 1989).  During a full-scale trial at the ITOSS (In Situ Treatment of Sediment) program in 1997 several remediation processes were tested on intermediate fuel oil (IF-30) artificially stranded on mixed (sand and pebble) intertidal shorelines.  The remediation methods included sediment relocation (surf washing), mixing (tilling) and bioremediation (Guenette et al. 2003; Sergy et al. 2003).  The fertilizers used included both soluble (prilled ammonium nitrate and superphosphate [Ca(H₂PO₄)₂]) and commercial slow-release (Inipol SP1) chemicals, applied to the top of the sediments during the first two months of the experiment.  The introduction of fertilizers resulted in elevated levels of bioavailable nitrogen and phosphorous in oiled sediments.  The biodegradation rates were approximately doubled over a period of one year in the oiled sediments that received fertilizers when compared to non-treated oiled sediments, and no acute toxicity was associated with the bioremediation treatment (Prince et al. 2003).  Mixing/tilling also seemed to result in increased microbial activity for limited periods by increasing sediment permeability (Owens et al. 2003).

Biostimulation field experiments have also been conducted in Antarctic environments. An Arabian crude oil was added to several 1 m² enclosures on intertidal sandy beaches on the main island of the Kerguelen Archipelago (49°19’S, 69°42.5’E).  Different fertilizers were added to the top of the oil, including the slow-release Inipol EAP 22 and various experimental mixtures consisting of dry fish compost, with or without supplements of urea, phosphate and charged or neutral surfactants (Pelletier et al. 2004).  During a 300-day experiment the oil was eventually depleted in both untreated and treated sediments in this cold environment (seawater temperatures 3-4°C), but the various fertilizers accelerated the biodegradation rates.  It was also observed that a fertilizer with a neutral surfactant reduced the toxicity of the oil during the last three months of the experiment (Delille et al. 2002).

Bioremediation was used as an oiled-beach cleaning technology on a full-scale oil spill in Arctic environments during the Exxon Valdez accident in March 1989.  This spill in Prince William Sound, Alaska, resulted in the release of 41 million liters of Alaskan North Slope crude oil.  Bioremediation was used extensively, employing the fertilizers Inipol EAP 22 and Customblen (slow-release granulated fertilizer).  Approximately 50,000 kg nitrogen and 5,000 kg phosphorous were applied to the shorelines over the summers of 1989-1992 (Bragg et al. 1994).  For a low-energy beach containing both surface and subsurface oil and treated with both fertilizers, it was estimated that the fertilizers enhanced oil biodegradation by 5.5 times over non-treated controls (Bragg et al. 1994).

Results from this bioremediation field experiment, supplemented with several laboratory experiments have been summarized by Pritchard et al. (1992):

• Fertilizers with slow release nutrients were recommended in tidal zones to avoid rapid wash away of the fertilizer components.  Solid granulated fertilizers were easier to apply over large areas, and adhered well to the oiled beach material. These fertilizers persisted over periods of 2-3 weeks after application.  Also fertilizer briquettes were tested with outcomes similar to the granules.  Burial of fertilizer material parallel to the water line has also been suggested.
• Components of the nitrogen and phosphorous from liquid Inipol EAP 22 seemed to be released rapidly when submerged in seawater.  However, residual nutrients remained and were available for oil-degrading bacteria.  Laboratory experiments indicated that the nitrogen part of the oleophilic fertilizer was more important than the phosphorous compounds for biodegradation.
• Inspection of areas treated with Inipol EAP 22 showed faster visual disappearance of oil from cobble surfaces than untreated surfaces, and laboratory experiments indicated biodegradation as the cause for disappearance rather than washing.  However, oil under the cobbles remained. It was also suggested that Inipol could work as a surfactant by chemically removing aliphatic hydrocarbons from samples.
• Measurements of bioremediation effectiveness should include analyses of several hydrocarbon groups, in addition to the aliphatic hydrocarbons.  Measuring compositional changes in the aromatic fraction by GCMS was therefore added as a further dimension to biodegradation.  Bioremediation should also result in removal of oil residues, measured as a reduction in total mass of the oil. 
• The use of common indicators of biodegradation like nC18/Phytane was problematic since phytane in some instances disappeared at the same rate as nC18 alkane (phytane-degrading microbes were actually isolated from beach material).  For aromatics a number of methyl-substituted homologues close in mass number to their parent structures were selected, and these were normalized to hopane, which are resistant to biodegradation.
• Results from oil compound analyses showed that reduction in nC18/Phytane ratios resulted in corresponding changes in aromatic and heterocyclic hydrocarbons.  Samples with reduced nC18/Phytane ratios also showed increased concentrations of compounds associated with asphaltenes and polar materials, which may be partially biodegraded material.  Further, there was a positive correlation between nC18/Phytane reduction and removal of residual oil, determined by mass.  It was therefore suggested that reduction in oil residues could be used as a measure of biodegradation.

Interactions between stranded oil and mineral fines have been used to reduce oil adhesion to solid surfaces and generate stable oil droplets released to the water column.  In this way the increased surface area of oil became more accessible to nutrients, oxygen and bacteria, resulting in increased microbial activity and oil biodegradation (Lee et al. 1996).

Efforts to stimulate crude oil biodegradation in seawater and ice (see next section) have not been investigated to the same extent as for stranded oil.  Most remediation strategies in oiled seawater have focused on mechanical removal methods such as the use of oil booms and skimmers.  In addition to mechanical methods, chemical dispersants have also been widely used as an alternative treatment. Dispersants are mixtures of surface-active chemicals that reduce the surface tension of the oil, resulting in the formation of oil droplets that are smaller than those initially generated by mechanical wave action (Brandvik 1997). These chemicals are primarily used to disperse oil spilled on the seawater surface into the water column.  This approach aims to reduce the impacts of oil spills on seabird and mammal populations in the vicinity of the spill and help prevent the oil reaching the coastline.  In addition, this process increases the oil-water interface and generates more bioavailable surface area for microorganisms.

The efficiencies of chemical dispersants on crude oil degradation at low seawater temperatures have shown conflicting results. In a seawater mesocosm experiment (3.5 m³ flow-through tanks with natural seawater from the St. Lawrence Estuary, Quebec) with temperatures of ‑1.8 to 5.5°C, Forties and Western Sweet Blend crude oils were treated with the dispersant Corexit® 9527 or different surfactant mixtures.  Over a 63-day period at water temperatures >0°C, chemical dispersal was found to result in higher biodegradation rates than in untreated oil samples (Siron et al. 1995).  In microcosm studies with a seawater temperature of 8°C, MacNaughton et al. (2003) added the dispersant Corexit 9500 to Alaskan North Slope (ANS) crude oil.  The dispersant resulted in rapid colonization of oil droplets by bacteria, and heterotrophic and oil-degrading microbes proliferated in the microcosms.  However, when the total hydrocarbons, C₁₁ to C₃₅ n-alkanes, or the sum of selected aromatics were assessed over a degradation period of 35 days, the addition of dispersant resulted in only slow or negligible biodegradation of the oil when compared to naturally dispersed oil.  In another biodegradation experiment, Lindstrom and Braddock (2002) exposed cultures of oil-degrading microbes to ANS (fresh or evaporated and spiked with radiolabelled hydrocarbons) dispersed with Corexit® 9500 over a 2 month period at 8°C.  Respirometric analyses were conducted to determine ¹⁴CO₂ mineralization of ¹⁴C-labelled dodecane, hexadecane, 2-methyl-naphthalene and phenanthrene. The dispersant was found to inhibit degradation of some of the hydrocarbons (hexadecane and phenanthrene), while others (dodecane and 2-methyl-naphthalene) were unaffected when compared to mineralization of the oil without dispersant.  It was suggested that carbon mineralization, at least initially, was the result of dispersant mineralization rather than degradation of the oil compounds. In a biodegradation study comparing two dispersants, Corexit® 9500 and JD2000, fresh Prudhoe Bay crude oil was mixed with the dispersants and a microbial culture originating from the shorelines contaminated during the Exxon Valdez accident.  Biodegradation of n-alkanes and PAHs was measured over a period of 46 days at 5°C.  The first-order degradation rates of most n-alkanes and PAHs were found to be higher with dispersants than without dispersants, although these data were not statistically significant.  When the same experiment was performed at 20°C the degradation rates of n-alkanes and PAHs were more rapid than at 5°C, showing temperature-related biodegradation rates both in non-dispersed and dispersed oil. Furthermore, the influence of the two dispersants on biodegradation differed between the two degradation temperatures (Venosa and Holder 2007).  Correlation studies between the droplet surface area of dispersed oil and the resulting degradation indicated that both dispersed area and dispersant chemistry controlled the degradation and that the surfactant blend hydrophile-lipophile balance and treatment levels were also significant controlling factors (Varadaraj et al. 1995).  Recently, Prince and Butler (2013) suggested that the variability in results may result from low oil concentrations in a physically dynamic experimental system which would act to form only slightly larger oil droplets than dispersed oil treatment.  Additionally, the authors indicate that an accelerated rate of biodegradation occurs when a dispersed oil treatment is compared to biodegradation of a surfaced oil, principally due to relatively small droplet size.

A variety of bacteria and yeasts also produce biosurfactants like rhamonolipids, sophorolipids and surfactin.  Biosurfactants consists of fatty acid hydrophobic parts and carbohydrate, amino acid, cyclic peptide, phosphate, carboxylic acid or alcohol as hydrophilic part (Mulligan 2005). Most biosurfactants are produced from hydrocarbon substrates (Syldatk and Wagner 1987).  It has been suggested that rhamnolipis addition can enhance biodegradation of hydrocarbon mixtures in liquid systems and soil (Maier and Soberon-Chevez 2000).  Two mechanisms for enhanced biodegradation were proposed; enhanced substrate solubility and interactions with the cell surface to increase the hydrophobicity of the cell surface, allowing improving association of hydrophobic hydrocarbons (Mulligan 2005).  Combined use of rhamnolipids and slow-release fertilizers (Inipol EAp-22) also enhanced biodegradation of aromatic and aliphatic hydrocarbons in liquid phase and soil (Churchill et al. 1995).  Studies withRhodococcus sp. Q15 grown on hexadecane or diesel fuel at 5°C showed production of biosurfactants at low temperature (5°C), indicating that the cell surfaces became more hydrophobic  (Whyte et al.1999).

In addition to chemical dispersants use of fertilizers to increase oil compound biodegradation has been investigated to some extent.  As early as in the 1970s enhanced biodegradation of oil spills by lipophilic slow-release fertilizers was investigated (Olivieri et al. 1975).  In a study with hydrocarbon-degrading bacteria (Alcanivorax sp.) and an open seawater-based system at 30°C guano was used as fertilizer.  It was demonstrated that commercial guano was an effective source of nitrogen and phosphorus for the growth of bacteria on crude oil, and that the guano resulted in extensive biodegradation of crude oil (n-alkanes (C10- C36) and polyaromatics at (Knezevich et al. 2007).

Bioremediation of oil in ice is an intriguing prospect. If biodegradation of crude oil could be stimulated in ice, especially for the most toxic compounds migrating out of the ice through the brine channels, this would be of benefit for organisms inhabiting the polluted ice or nearby areas.   Studies have shown that fertilizers can stimulate biodegradation of crude oils in cold seawater under controlled experimental conditions (Delille et al. 1998). The slow-release oleophilic fertilizer Inipol EAP 22 was added to Antarctic seawater contaminated with “Arabian light” crude oil in a mesocosm study.  The experiment was completed over 5 weeks during the Austral summers of 1992/1993 and 1993/1994. In both ice-covered and ice-free seawater, the addition of the fertilizer enhanced both the concentrations of heterotrophic and hydrocarbon-degrading bacteria and increased the rate of biodegradation during the experiments, measured as n-C₁₇/Pristane and n-C₁₈/Phytane ratios.

A winter field experiment was conducted at Svalbard in 2004 as part of the Arctic Operational Platform (ARCOP) program.  Crude Statfjord oil with and without fertilizers (mixture of Inipol EAP 22 and fish meal) was placed in fjord ice (Van Mijen Fjord, Svea) for a period of 6 months (December 2004-June 2005).  At sub-zero temperatures no significant degradation of oil hydrocarbons occurred with the addition of nutrients, but at 0°C melt pool samples fertilized with inorganic nutrients showed a significant change in bacterial diversity (Gerdes and Dieckmann 2006).  Importantly, many of the available slow-release fertilizers are not suitable for use in Arctic regions as they will solidify if used in ice at very low temperatures. For example, the pour point of Inipol EAP 22 is 11°C which makes it difficult to use effectively under Arctic conditions.  As a result slow-release fertilizers will require reformulation or new products will need to be developed specifically for use at very low temperatures.

A few studies have also been conducted to determine the impacts on ice protist communities after oil contamination and subsequent fertilizer treatment.  One of these studies formed part of the ARCOP field trial on Svalbard.  In oil-contaminated ice (no fertilizers) the protist communities were destroyed through complete ice coverage.  Upon addition of fertilizers a less pronounced decline of organisms in the ice interior was observed. Thus, the use of fertilizers (Inipol and fish meal) helped to maintain higher diversity and biomass of protists in the ice. In a separate study, heterotrophic flagellates appeared to escape or avoid the oil contamination by downward migration (Ikävelko et al. 2005).  In an Antarctic field experiment conducted during the Austral winter of 1993, land-fast ice on the continental shelf of Terre Adélie was contaminated with crude oil (Arabian light) or diesel fuel, and negative effects on the ice microalgae were determined by chlorophyll A measurements.  In crude oil-contaminated ice, negative effects were induced which lasted throughout the ice coverage period.  The diesel contamination studies were found to cause an even more rapid effect on the algae than the crude oil. However, the addition of the fertilizer Inipol EAP 22 resulted in clearly favorable effects on the sea ice microalgae (Fiala and Delille 1999).

Bioaugmentation has been proposed as a bioremediation method for soil and sediments, often as a supplement to biostimulation treatments. Introduction of exogenous hydrocarbonoclastic bacteria for detoxification of hydrocarbon-polluted cold environments has been reported, with variable success. A number of commercial products exist, which include microbial inocula.  These products are often lacking essential information about the bacterial content.  National authorities may also be skeptical about using these products without proper product information.

In a study of diesel oil-contaminated Alpine soil, a psychrophilic diesel oil-degrading inoculum was added to the contaminated soil, but biostimulation with fertilizers proved more efficient than the bioaugmentation for improved biodegradation activity (Margesin and Schinner 1997).  In microcosm experiments performed in Antarctic gas-oil polluted soil (Jubany Station, King George Island, South Shetland Islands) inoculation of the psychrotolerant strain B-2-2 resulted in 75% hydrocarbon removal, whereas 35% hydrocarbon removal was observed by biostimulation methods when compared to abiotic controls (Ruberto et al. 2003).

Several bioaugmentation studies from marine environments have been reported, although none of these are from cold waters.  A laboratory biodegradation and toxicity study of 12 commercially available bioaugmentation products applied to weathered oil (Alaska North Slope) in seawater at 20°C showed that 3 of the products enhanced biodegradation more than nutrient-amended controls, but only one product resulted in reduced toxicity (Aldrett et al. 1997).  In a marine sediment microcosm study the aromatic-degrading bacterial strain Cycloclasticus sp. E2 was shown to play an important role during degradation of naphthalene in combination with biostimulation treatment (Miyasaka et al. 2006).  Interestingly, bacteria from this genus were also abundant during bioremediation treatment of Arctic oiled beaches at Svalbard (Grossman et al. 2000).

Bioaugmentation has often proved inferior to biostimulation.  One plausible explanation for this may be that the introduced bacteria will have an immediate effect due to the biomass added, but these exogenous microbes may gradually be outcompeted by the indigenous microbes adapted to the local environment.   In a study with small- and mesoscale systems, addition of inorganic nutrients was more efficient at enhancing oil biodegradation in sediments than a commercial product consisting of nutrients and bacterial inocula.  The use of the bioaugmentation product suppressed both the rate and the extent of oil loss by tidal activity and biodegradation, when compared to the periodic addition of inorganic nutrients (Lee et al. 1997).

Purposeful acceleration of biodegradation through, for example, ecological engineering requires an understanding of the microorganisms that are present in the local environment and their responses to stimuli such as those outlined in the previous sections.  Advances have been made in this area, including development of techniques and submersible vehicles that allow evaluation of biodegradation in situ, improvements in laboratory equipment to simulate the natural environment, maturation of molecular methods to track population dynamics during the degradation process, and development of techniques that allow better understanding of functions carried out by those populations.  These topics were dealt with in more detail in the analytical methods subsection of this section entitle Determining Biodegradation section.  The –omics approaches are of particular importance to accelerated biodegradation.  The effects of management practices on microbial populations and functions, with resulting effects on the components of degraded oil and rates of degradation, can begin to be understood through the emerging fields of metagenomics, metaproteomics, and metabolomics.  These techniques promise to provide much greater understanding of microbial responses, which in turn will allow evaluation and management of oil degradation processes in the Arctic environment. 

The review of the biodegradation of oil in the Arctic described by the authors in this section led to suggestions of further research which can reduce remaining uncertainties. The more generic suggestions compiled from this review are summarized below while recommendations that are important for improving Arctic NEBA are listed separately. 

1. Determine effects oil properties may have on biodegradation under Arctic conditions. The physico-chemical characteristics and weathering conditions of different oils at low temperatures may vary considerably, having impacts on biodegradation efficiencies and should be considered for further research, for instance, in the relationship between biodegradation and oil appearance (e.g. viscosity, dispersibility, resurfacing) after a spill. Temperature-related biodegradation data used in models, often based on Q10-approaches, may show erroneous results at very low temperatures, probably due to physical changes in the oil. 
2. Further document the rates of weathering of stranded oil. Oil residuals from treated or untreated oil undergo different rates of weathering when transferred to sandy beaches or cobble beaches, under or within annual or multi-year ice and within lagoons or near estuaries. Stranded oil has a different potential for release of oil from those compartments and thus longer or shorter periods of time where localized re-oiling of adjacent locations occurs. 
3. Explore options of enhanced bioremediation with biostimulation and/or bioaugmentation. Biostimulation at very low temperatures where hydrocarbon biodegradation is apparent should be a research issue. Accelerated biodegradation has mainly been investigated by adding fertilizer formulations to improve the oil biodegradation capabilities of natural microbial populations. Biostimulation is regarded as a relevant tool for removal of stranded oil residues as an extension of mechanical treatment. Several formulations of fertilizers have been tested experimentally and under real spill situations. Oleophilic fertilizers have proven more efficient than inorganic nutrients, but available formulations seem to have reduced effects at very low temperatures, such as in marine ice (Gerdes and Dieckmann 2006). Bioaugmentation seems to be less efficient than biostimulation, although some experimental studies have combined the two remediation options. However, several oil cleanup products have been marketed in recent years, including bacteria. Further development of techniques for the enrichment of site-specific oil-degrading communities may therefore be of relevance to combine with other remediation techniques. 

The recommendations presented below indicate where increased knowledge of biodegradation processes would result in reducing existing uncertainties in NEBA assessments.  No prioritization has been made to the list; for some of the recommendations, surrogate data may be already available.

1. Investigate potential for recalcitrant effects of oil and consequences of oil spill response strategies (OSRs).    There is a need to better understand what happens to the unrecovered oil once released and distributed to different Environmental Compartments (ECs) in Arctic environments, such as surface slicks, and under ice or onshore stranding.  In order to meet these objectives, we recommend conducting a series of laboratory experiments to determine the persistence, rate and extent of biodegradation in relevant Arctic systems, including surface water, ice and sediment under Arctic marine conditions to further our understanding of the biodegradable and recalcitrant oil dynamics in these compartments.
2. Examine the influence of ice on biodegradation rates.  Marine ice poses a particular challenge to the indigenous microbes, which require the ability to survive and be metabolically active at sub-zero temperatures and at high salt concentrations.  Microbial metabolism has been demonstrated in marine ice, and populations have been stimulated by oil in the ice.  However, the extent of hydrocarbon biodegradation in Arctic ice requires more attention.
3. Improve knowledge of degradation rates of dispersed oil and oil in the water column and in other Arctic environmental compartments.  More information is needed on how employment of different oil spill response strategies would affect biodegradation rates and oil movement to different compartments.  For investigations of OSR strategies, we suggest studies be conducted with chemically dispersed oil. These studies include experiments relevant for the following compartments and conditions:
   1. The seawater surface with droplets entering the surface creating a surface film/sheen
   2. The water column, creating physically or chemically generated dispersions and maintaining these in the water column
   3. Subtidal seabed, examining the biodegradation of oil remaining on the seabed after particle integration
   4. Oil in marine ice, with dissolution of water-soluble components to the brine channels in the ice
   5. Shoreline sediments, examining the biodegradation of oil stranded in areas that are ‘primed’ by prior exposure to oil compared with biodegradation in areas that are ‘naïve’ or previously unexposed to oil

Authors Dr. Odd Brakstad (SINTEF), Dr. Donald Stoeckel (Battelle), 

Dr. Charles Greer (NRC Canada)