Oil reservoirs, an exceptional habitat for microorganisms.

Microorganisms are present in oil reservoirs around the world where they degrade oil and lead to changes in oil quality. Unfortunately, our knowledge about processes in deep oil reservoirs is limited due to the lack of undisturbed samples. In this review, we discuss the distribution of microorganisms at the oil-water transition zone as well as in water saturated parts of the oil leg and their possible physiological adaptations to abiotic and biotic ecological factors such as temperature, salinity and viruses. We show the importance of studying the water phase within the oil, because small water inclusions and pockets within the oil leg provide an exceptional habitat for microorganisms within a natural oil reservoir and concurrently enlarge the zone of oil biodegradation. Environmental factors such as temperature and salinity control oil biodegradation. Temperature determines the type of microorganisms which are able to inhabit the reservoir. Proteobacteria and Euryarchaeota, are ubiquitous in oil reservoirs over all temperature ranges, whereas some others are tied to specific temperatures. It is proposed that biofilm formation is the dominant way of life within oil reservoirs, enhancing nutrient uptake, syntrophic interactions and protection against environmental stress. Literature shows that viruses are abundant in oil reservoirs and the possible impact on microbial community composition due to control of microbial activity and function is discussed.

Introduction

Oil reservoirs are extreme environments for microbial life [1] characterized by high toxicity, hydrophobicity and low water activity, as well as high temperature, salinity, and pressure [2]. Nevertheless, oil reservoirs offer a broad range of niches for a multitude of bacteria and archaea, such as sulfate-, nitrate-, and iron-reducers, fermenters, acetogens, and methanogens [1,3]. The microbial degradation of oil results in a higher fraction of bitumen and eventually leads to the deterioration of the world’s oil resources. Since oil is still one of the most important resources for industry and energy [4], it is crucial to gain insights into the microbiology of oil reservoirs. Over the past decades, numerous reviews on oil microbiology have investigated the extent of biodegradation, the effect of microbes on oil quality, oil production methods and enhanced oil recovery (EOR) [3,[5], [6], [7], [8]], as well as their taxonomical and functional composition and the impact of environmental factors on microbes [3,7,9,10]. However, due to the lack of undisturbed samples, our knowledge of microbial ecology in oil reservoirs is still limited [11].

Distribution of microbes within oil reservoirs

Oil reservoirs consist of different phases where microorganisms can thrive, such as crude oil, formation water and solid surfaces from rock and organic materials [12]. To understand the oil-water distribution patterns of microbes, it is important to conceptualize the oil habitat. In general, microbial degradation of oil is limited by the availability of electron acceptors because, due to thermodynamic constraints, hydrocarbons cannot be fermented without a hydrogen and acetate scavenging process. However, microbes can only conserve energy if they have direct contact to both electron donors from the oil phase and electron acceptors from the water phase [13,14]. This situation is found at the oil-water transition zone (OWTZ) beneath the oil leg, which is a hotspot of microbial growth and oil degradation [5,15] (Fig. 1). Here, the oil phase provides electron donors and the water phase provides the habitat for the microorganisms. Consequently, the rate of oil biodegradation depends strongly on the size of the surface of the oil-water interface.

Fig. 1

Schematic scheme of a deep subsurface oil reservoir (oil leg) with underlying brine water (water leg). Most of the biological oil degradation takes place at the oil-water transition zone (OWTZ) and in dispensed water droplets nearby. Microorganisms live attached to rock particles in a thin water film or in dispensed water droplets amidst the oil phase.

In deep oil reservoirs, dissolved electron acceptors oxygen and nitrate are naturally absent unless anthropogenically added via injected fluids [8]. Several studies detected iron-, manganese-, or nitrate-reducing bacteria such as Shewanella putrefaciens or Deferibacter thermophilus in fluids of oil reservoirs [[16], [17], [18], [19], [20]]. However, solid iron(III) and manganese (IV) oxides typically are not available as electron acceptors for microbial oil degradation because they have been reduced over the millions of years and are not replenished. Therefore, the most prevalent processes are fermentation, methanogenesis and sulfate reduction, if there is a source of sulfate [8,21,22].

It has been a paradigm for the last decades that biodegradation mostly takes place directly near the OWTZ. However, studies have shown an increasing saturated hydrocarbon content over approximately 100–130 m away from the OWTZ due to circulation and diffusion [5]. Nevertheless, there is always a certain amount of water also present in the oil leg either as water-saturated areas in the pore space of the rock or as a thin water film covering the rock surface in water wet reservoirs. Thus, different oil reservoirs vary in water content as well as in their oil composition [5,7,23,24]. Interestingly, oil samples with high water content of around 40–60% harbor a 2.6-fold higher bacterial richness compared to low water content oils with 1–5% [25] indicating that the amount of water present in the oil leg plays an important role as a habitat for the microorganisms.

In fact, stratified water pockets and pore spaces were discovered in the natural asphalts from the La Brea Tar Pits in Los Angeles, CA [26], and small water inclusions were discovered in the oil phase of the Pitch Lake in Trinidad and Tobago, the world largest natural tar lake [27]. The 1–3 μl water droplets from Trinidad and Tobago were densely populated with complex microbial communities and actively degrading the oil. The high salinity and water-stable isotope measurements indicated that the water droplets originated from deep subsurface formation water, most likely directly form the oil reservoir feeding the natural oil seep through a geological fracture [27]. Taking the water droplets as a proxy for subsurface processes, it is very likely that significant microbial populations can thrive within water-filled rock pores away from the OWTZ. Consequently, microbial degradation potential in the reservoir should correlate with the water content of the different phases, building a gradient of degradation activity starting from the OWTZ at the bottom and decreasing to the top of the reservoir (Fig. 1). In fact, such patterns of biodegraded oil can be found in reservoirs although they have been interpreted as diffusion gradients of alkanes from the non-degraded oil at the top of the reservoirs towards the depletion hot spot at the OWTZ [11]. Nevertheless, the findings from Tar Pits in Los Angeles and of the dispersed water droplets in the natural oil seep in Trinidad indicate that microbes reside in water pockets within the oil phase or even in the water film around sand grains [26,27]. This concept enlarges the overall oil-water interface and should be considered as having a large impact on the degradation process. Microbial oil biodegradation can consequently occur not only at the OWTZ, but also within the oil leg (Fig. 1).

Upon production, oil is pumped to the surface as an artificial mixture of water, oil and gas. One has to be aware that this mixture does not necessarily contain the true composition of the microbial communities in the subsurface. Most of the microbes that thrive either at the OWTZ or in the water-filled compartments in the oil leg will thrive attached to the rock matrix rather than planktonic in the water phase and will not appear in oil or produced water. Furthermore, cells not attached to the rock are probably present in dense biofilms at the oil-water interface.

Most studies on oil microbiomes classify samples from oil wells as either oil or water phase. The natural water phase mostly consists of formation water, which is naturally present within porous rock as explained above [3,28]. The amount and composition of formation water, its natural flow-rate and its flow-path through the oil field can influence the microbial community [3,24,29]. Most studies assume that DNA extracted from the formation water represents the microbiome of the total reservoir [28,30,31]. However, analysis of water- and oil phases have revealed vast differences between their microbial community compositions [13]. The bacterial diversity in the oil phase [28] and species richness [12] appeared to be much higher than those of the water phase. Some bacteria might attach directly to the oil surface, while others remain in the water phase. Acidobacteria, Actinobacteria, Fusobacteria, Nitrospira, Pseudomonadales and Thermodesulfobacteria are more frequently isolated from the oil phase, whereas Alphaproteobacteria, Atribacteria, Bacteriodetes, Betaproteobacteria, Campylobacterales, Chloroflexi, Synergistetes and Thermotogae are mostly isolated from the water phase [9,28].

Thus, the water phase itself contains only a minor portion of the microorganisms in the oil reservoir and can only constitute a proxy for the true community composition. The oil phase contains another part of the microbial communities, most likely containing attached microbes and those present in small water droplets dispersed in the oil [9,27]. Thus, at least the two phases, water and oil, should be analyzed to obtain a better picture of the true microbial community composition.

Life in extreme environments – oil as an exceptional habitat

Biofilms as a physiological adaptation to life in oil?

Microbial life in oil reservoirs faces severe conditions with multiple stressors such as toxicity of the oil and low water activity. In addition, planktonic microorganisms often do not have access to both electron donor and electron acceptor as a prerequisite for microbial energy conservation and must attach to the oil-water interface. One microbial adaption to these harsh conditions is organization in biofilms. Microbial biofilms are widely distributed in nature and belong to the most successful strategies of life on earth [32]. The formation of biofilms was shown in the presence of crude oil under aerobic conditions such as in an oil lake in the Kuwaiti desert [33] or as the formation of microbial mats by cyanobacteria in seawater [34,35]. Generally, biofilms are defined as ‘aggregates of microorganisms in which cells are frequently embedded in a self-produced matrix of extracellular polymeric substances (EPS) that are adherent to each other and/or a surface’ [36]. EPS are a gel-like network that keep the microbial aggregates together in order to provide mechanical stability to biofilms in flowing or moving systems such as water and oil [[37], [38], [39], [40], [41]]. The sticky coating of a biofilm matrix is a protective layer against external environmental stressors including desiccation, temperature and salt concentrations in oil fields. In the presence of crude oil, many bacteria produce EPS, which act as biosurfactants to enhance the solubility and bioavailability of hydrophobic organic compounds, such as polycyclic aromatic hydrocarbons (PAH) or n-alkanes [[42], [43], [44], [45], [46], [47], [48]]. In addition, a biofilm provides space for mutualistic micro-consortia, confined gradients increasing habitat diversity, resource capture by nutrient sorption, nutrient transport by channel systems, exchange of signaling molecules and acts as a barrier against toxic compounds by increased tolerance due to horizontal gene transfer [[49], [50], [51], [52]]. Because of the protective and nutrient transport properties of biofilms, it is not surprising that they have been found at the alkane-water interface of n-alkanes (C8-C28) and n-alcohols (C12 and C16) [45,47,53]. Some enrichment cultures were able to build biofilms on the surface of phenanthrene and other PAHs to overcome the mass transfer limitations during the degradation [44,46]. Biofilm formation also depends on the solubility of the PAHs; lower solubility results in more attached cells and biofilm formation to overcome the mass transfer limitations [46]. Due to these advantages of biofilms in toxic and extreme environments, we propose that biofilm formation is also the predominant form of life in oil reservoirs.

Metabolic functions and interactions of microorganisms in oil reservoirs

In the absence of the most favorable electron acceptors oxygen and nitrate, sulfate reduction and syntrophic methanogenesis are the dominant processes in oil biodegradation [3]. If sulfate is present at concentrations higher than 50 μM, hydrocarbon degradation coupled to sulfate reduction is the dominating process over methanogenesis [54]. Sulfate-reducing microorganisms are phylogenetically diverse and can be found within the Proteobacteria, Firmicutes, Nitrospira and Thermodesulfobacteria, as well as in the Crenarchaeota and Euryarchaeota [55]. If sulfate is absent as electron acceptor, many sulfate-reducers can switch their metabolism to fermentative oil degradation, producing short chain fatty acids, molecular hydrogen and carbon dioxide [56,57]. For example, members of the genera Desulfovibrio, Desulfotomaculum and Archaeoglobus can grow with sulfate as electron acceptor or as fermenters in association with methanogens when sulfate is depleted [56,[58], [59], [60], [61]]. Fermentation of hydrocarbons is thermodynamically only feasible when coupled to methanogenesis depleting both hydrogen and acetate [22,59,[62], [63], [64], [65], [66], [67], [68]]. Thus, methanogenic oil degradation is always a syntrophic process, where different members of the microbial community perform different steps in an overall metabolic process which cannot be fulfilled by a single member alone [69]. In addition, methane can be generated by acetoclastic methanogens disproportioning acetate to CO2 and methane or by methylotrophic methanogenesis [70]. In fact, most of the biological methane generated in oil reservoirs originates from acetoclastic methanogenesis as indicated by stable isotope signatures [71]. Syntrophic interactions in oil reservoirs are not confined to a specific phylogenetic group of prokaryotes [56,59,63,72]. The two bacterial phyla Proteobacteria and Firmicutes and the three classes Archaeoglobi, Methanomicrobia and Methanobacteria affiliated to Euryarchaeota, are mostly involved in syntrophic interactions (Table 1) [56,63,[72], [73], [74], [75]]. Many Proteobacteria are known to be syntrophic alkane degraders, e.g. members of the genera Marinobacter and Smithella have been highly enriched in methanogenic oil-degrading cultures [63,68,72,73,76]. Members of the genus Syntrophus can degrade alkanes and fatty acids in syntrophic association with methanogens [64,72]. As methanogenic and fermentative microorganisms are strongly dependent on each other, they are frequently organized in close vicinity to each other in order to provide a rapid exchange of electrons by diffusion of hydrogen or formate [67]. Furthermore, an electron exchange by direct interspecies electron transfer has been discovered; Geobacter metallireducens transfers electrons directly to Metanosaeta harundinacea during methanogenic degradation of ethanol, presumably by nanowires [77,78]. In addition, inorganic, electrically conductive particles inside a biofilm matrix can support interspecies electron transfer [79]. Thus, we suggest that the methanogenic degradation of oil mainly takes place in mutualistic microbial consortia organized in biofilms, where fermenting microbes transfer electrons either directly or indirectly to the methanogens.

Table 1

Phylogenetic affiliation of syntrophic consortia in oil reservoirs and their metabolic potentials.

phylum / class	order / genus	metabolic capacities	references
Proteobacteria / Gammaproteobacteria	Alteromonadales / Marinobacter	syntrophic alkane degraders	[63]
Proteobacteria / Deltaproteobacteria	Syntrophobacterales / Smithella
Proteobacteria / Deltaproteobacteria	Syntrophobacterales / Syntrophus	syntrophic fatty acids and alkanes degraders in association with methanogens	[64]
Proteobacteria / Deltaproteobacteria	Desulfovibrionales / Desulfovibrio	metabolically versatile, sulfate respiration and syntrophic alkane fermenters in association with methanogens	[56,60]
Firmicutes / Clostridia	Clostridiales / Desulfotomaculum		[56,61]
Firmicutes / Clostridia	Clostridiales / Clostridium		[56,73]
Euryarchaeota / Archaeoglobi	Archaeoglobales / Archaeoglobus	sulfate-reducing archaea and syntrophic fermentative alkanes degrader in association with acetotrophic methanogens	[59,72]
Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanoculleus	hydrogenotrophic methanogens	[56,73]
Euryarchaeota / Methanomicrobia	Methanosarcinales / Methanosaeta	acetotrophic methanogens	[56,73]
Euryarchaeota / Methanobacteria	Methanobacteriales / Methanobacterium	hydrogenotrophic methanogens	[72]

Anthropogenic impacts – injection water

Oil production is the major anthropogenic factor influencing microbial communities in oil reservoirs. This includes drilling, flooding, hot steam and water injections, all of which lead to a high potential of invasion of external microorganisms into the original microbial communities [9,24,29]. Water injections are necessary in secondary oil production stages to increase reservoir pressure. The amount of injected water depends on the reservoir pressure, well age or water progressing within the reservoir. Oil companies use different types of injection waters consisting of either seawater, fresh water or recycled formation water. Offshore fields are mostly supplied with seawater [24,80] whereas in other oil fields groundwater [81] or surface water [30] are used. In some reservoirs, injection waters are enriched with chemicals or nutrients in order to manipulate the indigenous microbial community. For instance, nitrate and nitrite injections are used to suppress reservoir souring by microbial H2S production. Oxygen injections can stimulate the aerobic hydrocarbon metabolism and mobilize the oil within the well by lowering the interfacial tension between oil and water phase through biosurfactant producing microbes or changes in the oil matrix [82]. Alternatively, fermentative bacteria and carbohydrate injection lead to the generation of acids, gases and solvents, which increases oil output, so called enhanced oil recovery (EOR) [8,31]. Water injections decrease the temperature of the oil field and build up a temperature gradient. The injection volume also affects the chemical composition of the production water, as it decreases the concentrations of magnesium, potassium, nitrate, nitrite and sulfate [24]. Production water is a byproduct of oil production and has been transported through the oil phase and pumped to the surface; it can be a mixture of formation and injection water and can contain particles and soluble compounds from oil [11,83] (Fig. 1). Two comparative studies have demonstrated higher concentrations of ammonium and fatty acids in formation water compared to production water [24]. Produced fluids with less than 10% injection water content did not have significant influence on microbial composition and metabolic potential. In contrast, fluids with a higher injection water cut indicated that the community composition and metabolic potential can be altered by the water composition. A close correlation was calculated between the relative abundance of the genus Flexistipes, family Deferribacteres and the proportion of injected seawater and the concentrations of magnesium, potassium, nitrate, nitrite, and sulfate. Epsilonproteobacteria and Gammaproteobacteria were isolated in greater abundance from sample wells with the highest water injection rate [24]. Several studies have reported a relationship between the chemical composition of the oil reservoirs and the operational taxonomic units (OTUs) found therein. In Algerian oil reservoirs, production waters revealed significant correlations between the relative abundance of bacterial OTUs or phyla and Cl− and K+ ions. However, it is not clear if this correlation is causative. It may well be that the real causes for microbial community differences are for example differences in water content and structure of the reservoir and that the ion composition is just an indication of geological differences. Significant differences between microbial community composition of production and injection waters were observed for waters from Algerian oilfields and the offshore Halfdan oil field in the Danish North Sea. The Algerian oilfield injection water was richer in cells and dominated by bacteria, whereas the production water contained ten times fewer cells and was dominated by Archaea [24,81]. This difference indicates a trivial correlation between oil degradation processes in the methanogenesis-dominated reservoir and the microbial community composition in the production water.

Oil quality is determined by the degree of biodegradation and physical processes during oil production such as water injection or phase fractionation explained above. Those processes lead to lower concentrations of specific isomers, hydrocarbons, sulfur-, oxygen- and nitrogen-containing compounds [84,85] and an increase in oil viscosity, metals and microbial metabolites such as organic acids or sulfur compounds. Biodegraded oil reservoirs typically consist of oil-water emulsified fluids and systematic gradients built by different oil components [8,24]. For instance, isoprenoids and n-alkanes concentration decrease towards the OWTZ as they are degraded faster than aromatic compounds [11,86]. There, the degradation of oil is also controlled by the nutrient availability in the individual reservoir [8]. In contrast to many oil reservoirs, the bulk water contained in oil of the Pitch Lake in Trinidad and Tobago revealed that concentrations of essential nutrients, such as 95 mg/L ammonia and 5 mg/L phosphate, were not growth-limiting [27] demonstrating again that conditions within the oil phase allow biodegradation [87].

Abiotic factors – temperature, pH, and salinity

The geology of an oil reservoir determines the temperature, pH and salinity, which influence the composition and metabolic activity of the indigenous microbiota. Temperature is one of the most important factors determining microbial community composition in oil reservoirs [1,3,[8], [9], [10],24,29,81]. Temperature increases by about 2–3 °C per 100 m of depth, which means that the effects of depth and temperature are closely related [3,81]. Generally, temperature is higher in reservoirs during primary production before injection, compared to similar reservoirs where water injections cool down the reservoir during secondary production [9,24]. The maximum temperature for hydrocarbon degradation in oil reservoirs is generally accepted to be around 82 °C [3,88]. The extreme solvent stress of the oil increases with elevated temperature and most likely the integrity of the cell membranes suffers. A study detected hyperthermophilic microorganisms in reservoirs with well temperatures up to 131 °C [80]. However, as the real conditions in the habitat could not be determined, it is highly unlikely that the organisms really thrived at that temperature in situ. So far, the record in hyperthermophilic growth is at 95 °C by the bacterium Aquifex pyrophilus and at 113 °C by the archaeon Pyrolobus fumarii, which of course did not take place in the presence of hydrocarbons [89]. Therefore, the detection of life at 131 °C is questioned by the indirect estimation of the oil temperatures and the so far known temperature maxima of microbes [10,80,81,90].

Nevertheless, highest microbial diversity has been found at moderately hot reservoirs with temperature of around 55 °C [91]. As everywhere in the environment, oil reservoirs harbor microorganisms with different temperature preferences (Table 2). Nitrospira, Atribacteria and Acidobacteria were only detected in high-temperature oil reservoirs above 50 °C. Most Gammaproteobacteria, like Firmicutes, Thermotogae and Thermodesulfobacteria, showed a higher relative abundance in high-temperature oil reservoirs above 50 °C. Spirochaetes, Synergistetes, Chloroflexi, Marinobacterium, Paracoccus, Donghicola and Planctomycetes were more frequently detected in oil reservoirs below 50 °C. The archaea Haloferacales, Thermoproteales, Sulfolobales, Nitrososphaerales, Halobacteriales, Fervidicoccales and Thermoplasmatales have been detected exclusively in high-temperature oil fields above 50 °C. Thermococcales and Archaeoglobales are known as thermophilic lineages and have frequently been isolated from oil reservoirs above 70 °C [24,29,92]. Methanobacteriales (e.g. Methanothermobacter), Thermococcales (e.g. Thermococcus), Methanococclaesk and Archaeoglobales were most abundant in high-temperature oil fields [9]. Methanosarcinales, Methanomicrobiales (e.g. Methanocorpusculum and Methanolinea), Desulfurococcales and Methanocellales were mostly isolated from oil reservoirs below 50 °C.

Table 2

Bacteria and archaea typically associated with low-temperature (< 50 °C) or high-temperature (≥ 50 °C) petroleum reservoirs.

temperature optimum	phylum / class	order / genus	reference
ubiquitous	Proteobacteria / Epsilonproteobacteria	Campylobacterales / Arcobacter	[29]
	Proteobacteria / Epsilonproteobacteria	Campylobacterales / Sulfurospirillum
	Proteobacteria / Gammaproteobacteria	Pseudomonadales / Pseudomonas
	Proteobacteria / Alphaproteobacteria	Rhizobiales / Rhizobium
	Proteobacteria / Alphaproteobacteria	Sphingomonadales / Sphingomonas
	Acinetobacter
only > 50 °C	Crenarchaeota / Thermoprotei	Fervidicoccales	[9]
	Euryarchaeota / Halobacteria	Halobacteriales
	Euryarchaeota / Halobacteria	Haloferacales
	Thaumarchaeota / Nitrososphaeria	Nitrososphaerales
	Nitrospirae / Nitrospira	Nitrospirales / Nitrospira
	Crenarchaeota / Thermoprotei	Sulfolobales
	Proteobacteria / Deltaproteobacteria	Syntrophobacterales / Thermosulforhabdus	[93]
	Euryarchaeota / Thermoplasmata	Thermoplasmatales	[9,31]
	Crenarchaeota / Thermoprotei	Thermoproteales	[9]
	Acidobacteria		[9,29]
	Atribacteria		[9]
mostly > 50 °C	Euryarchaeota / Archaeoglobi	Archaeoglobales	[8,9,29]
	Firmicutes / Bacilli	Bacillales / Anaerobacillus	[29]
	Firmicutes / Bacilli	Bacillales / Bacillus
	Firmicutes / Clostridia	Clostridiales / Thermosyntropha
	Euryarchaeota / Halobacteria	Halobacteriales / Halogeometricum	[31]
	Proteobacteria / Hydrogenophilalia	Hydrogenophilales / Tepidiphilus	[29]
	Thermotogae / Thermotogae	Kosmotoga
	Euryarchaeota / Methanobacteria	Methanobacteriales / Methanothermobacter	[29,31,94,95]
	Euryarchaeota / Methanomicrobia	Methanocellales / Methanocella	[31]
	Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanocalculus	[29]
	Euryarchaeota / Methanomicrobia	Methanosarcinales / Methanosaeta	[29,31]
	Euryarchaeota / Methanomicrobia	Methanosarcinales / Methanomethylovorans	[31]
	Nitrospirae / Nitrospira	Nitrospirales / Thermodesulfovibrio	[29]
	Proteobacteria / Alphaproteobacteria	Rhodospirillales / Tistrella
	Deinococcus–Thermus / Deinococci	Thermales / Thermus
	Firmicutes / Clostridia	Thermoanaerobacterales / Thermoanaerobacter	[8,10]
	Euryarchaeota / Thermococci	Thermococcales	[8,9,24,29,31]
	Euryarchaeota / Thermoplasmata	Thermoplasmatales / Thermogymnomonas	[31]
	Actinobacteria / Thermoleophilia		[29]
	Bacteroidia / Bacteroidia
	Deferribacteres / Deferribacteres		[24]
	Firmicutes		[31]
	Proteobacteria / Betaproteobacteria		[29]
	Proteobacteria / Deltaproteobacteria
	Tenericutes / Mollicutes
	Thermodesulfobacteria		[31]
	Thermotogae
mostly < 50 °C	Actinobacteria / Actinobacteria	Actinomycetales / Microbacterium	[29]
	Actinobacteria / Actinobacteria	Actinomycetales / Dietzia
	Actinobacteria / Actinobacteria	Actinomycetales / Rhodococcus
	Proteobacteria / Gammaproteobacteria	Alteromonadales / Marinobacterium
	Crenarchaeota / Thermoprotei	Desulfurococcales	[9]
	Euryarchaeota / Methanobacteria	Methanobacteriales / Methanobacterium	[31,95,96]
	Euryarchaeota / Methanomicrobia	Methanocellales	[9]
	Euryarchaeota / Methanococci	Methanococcales / Methanococcus	[29]
	Euryarchaeota / Methanomicrobia	Methanomicrobiales	[9,95,96]
	Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanocorpusculum	[29]
	Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanoculleus	[29,31]
	Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanolinea	[29]
	Euryarchaeota / Methanomicrobia	Methanosarcinales	[9]
	Euryarchaeota / Methanomicrobia	Methanosarcinales / Methanolobus	[29]
	Proteobacteria / Alphaproteobacteria	Rhodobacterales / Donghicola
	Proteobacteria / Alphaproteobacteria	Rhodobacterales / Hyphomonas
	Proteobacteria / Alphaproteobacteria	Rhodobacterales / Paracoccus
	Bacteroidetes		[31,97]
	Chloroflexi		[9]
	Planctomycetes
	Proteobacteria		[31]
	Spirochaetes		[9]
	Synergistetes

Cai et al. investigated four production wells comprising a range of 1620 m–2470 m depth and 35.5 °C–69.0 °C [1]. They found an increasing relative abundance of genes related to the degradation of aromatic compounds (nahA, HBH, and pobA), carbon cycling and metabolism of other organic compounds with decreasing temperature and depth of oil-containing stratum. Stress response (heat shock), antibiotic resistance, and sulfur metabolism associated genes decreased with decreasing temperature [1].

Taken together, a correlation between functional gene occurrence and reservoir temperature has already been detected in nature. Oil degradation is highest at lower temperatures and reservoirs are more or less sterile at temperatures above 80 °C. Temperature is one of the major factors influencing microbial community composition and function in oil reservoirs producing trends on a genus level but not always on an order level. In general, it can be concluded that Proteobacteria (Alpha-, Gamma-, and Epsilonproteobacteria) and Euryarchaeota are ubiquitous in oil reservoirs across all temperature ranges. Sulfate-reducing bacteria (SRB) thrive from 4-85 °C [9]. Hence, no general predictions can be made on phylum or class level based on the oil reservoir temperature. On order and genus level, however, we can see clear temperature preferences as some orders, most of them archaea, were only isolated from high temperature reservoirs. Thus, they may serve as indicators for determination of in situ reservoir temperature. However, a general prediction of microbial community composition based on temperature alone is still not possible.

In oil reservoirs, salinity concentrations range from almost fresh- to salt-saturated water. Even though salinity and pH have been much less examined than temperature, they also have a high impact on microbial communities in oil by affecting growth and limiting bacterial activity. It was found that Clostridia correlated with low salinity of 3.8%, while Petrotoga and Desulfotomaculum species were mostly found in samples with a higher salinity of 7.2% [24]. It was suggested that hydrocarbon-degradation by Desulfotomaculum species may occur even under relatively high salinity conditions [24]. The amount of microbes isolated from oil fields decreased with increasing reservoir salinity above 10% [10]. Sulfate-reducing bacteria where found to resist wide ranges of salinity from 0 to 17% [9]. Manipulating the salinity of the injection water during oil-production to NaCl concentrations above 12% inhibits microbial H2S production [8]. The analysis of two different pits from the La Brea Tar Pits in Los Angeles indicated that site-specific differences in salinity were highly correlated with microbial community structures within the asphalt [26]. Salt concentrations in oil reservoirs affected methanogenic oil biodegradation as hydrogenotrophic methanogenesis from CO2 with H2 was only measured up to a salt concentration of 9%, in situ [86,98] (Table 3).

Table 3

Bacteria and Archaea associated to salinity and pH in petroleum reservoirs.

phylum / class	order / genus	prefered salinity and pH	reference
Euryarchaeota / Methanobacteria	Methanobacteriales / Methanothermobacter	acidic pH	[29]
Proteobacteria / Gammaproteobacteria	Pseudomonadales / Pseudomonas	acidic pH (5.5-7.6)
Proteobacteria / Betaproteobacteria		acidic pH (5.5-6.5)
Proteobacteria / Epsilonproteobacteria		acidic pH (5.5-6.5)
Proteobacteria / Gammaproteobacteria		acidic pH
Euryarchaeota / Archaeoglobi	Archaeoglobales / Archaeoglobus	alkaline pH
Proteobacteria / Deltaproteobacteria	Desulfuromonadales / Desulfuromonas	alkaline pH
Euryarchaeota / Methanococci	Methanococcales / Methanococcus	alkaline pH
Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanocorpusculum	alkaline pH
Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanocalculus	alkaline pH
Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanoculleus	alkaline pH
Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanolinea	alkaline pH
Euryarchaeota / Methanomicrobia	Methanosarcinales / Methanosaeta	alkaline pH
Euryarchaeota / Methanomicrobia	Methanosarcinales / Methanolobus	alkaline pH
Proteobacteria / Alphaproteobacteria	Rhodobacterales / Paracoccus	alkaline pH (7.0-8.2)
Actinobacteria		alkaline pH (7.0-8.0)
Proteobacteria / Alphaproteobacteria		alkaline pH (7.0-8.0)
Firmicutes / Clostridia	Clostridiales / Desulfotomaculum	higher salinity	[24]
Proteobacteria / Deltaproteobacteria	Desulfovibrionales / Desulfovermiculus halophilus	higher salinity	[94]
Firmicutes / Clostridia	Halanaerobiales / Haloanaerobium	higher salinity	[96,100]
Euryarchaeota / Methanococci	Methanococcales / Methanothermococcus	higher salinity	[29]
Thermotogae / Thermotogae	Petrotogales / Petrotoga	higher salinity	[24]
Euryarchaeota / Methanobacteria	Methanobacteriales / Methanobacterium	lower salinity	[93,97]
Euryarchaeota / Methanomicrobia	Methanomicrobiales / Methanoplanus	lower salinity	[101]

The in situ pH values of oil reservoirs typically range from 3 to 7 [3]. Sulfate-reducing bacteria where not only found to resist wide ranges of salinity but also a wide range of pH values, from 4 to 9.5 [9]. A site-specific correlation between pH and microbial community structures was detected for two different pits from the La Brea Tar Pits in Los Angeles [99]. A study across 22 geographically separated oil reservoirs in China showed that Alphaproteobacteria, Deltaproteobacteria and Actinobacteria were most abundant in neutral to alkaline reservoirs with pH values between 7.0 to 8.2. Pseudomonas correlated with decreasing pH value of formation brine in the range of 5.5 to 7.6. Gammaproteobacteria, Betaproteobacteria and Epsilonproteobacteria preferred even more acidic environments and were detected in reservoirs with pH values of 5.5 to 6.5 [29] (Table 3).

Viruses in oil reservoirs

Viruses are known to have a major impact on microbial communities and their ecology [1,[102], [103], [104], [105], [106], [107], [108]]. By lysing their bacterial hosts, bacteriophages cause the release and turnover of nutrients such as proteins and nucleic acids [105,107]. Lysis can also result in changes in the bacterial community composition known as the ’killing the winner’ hypothesis [105,106,109], meaning that host-specific predators (viruses) attack a bacterial population if the bacterial density increases over a certain threshold abundance. It thus prevents a species from emerging and maintains the coexistence of all species in the system [110]. ‘Killing the winner’ models predict that density- and frequency-dependent viral predation suppresses rapidly growing hosts, which leads to increasing host diversity [111]. Nevertheless, phages can also integrate into the host’s genome as prophages (lysogeny). The ‘piggyback-the-winner’ model predicts relationships between virus-like-particles and host densities. The main advantage for the phage lies in continuous proliferation by the regular host cell growth and division, without killing the host [111,112]. Prophages protect the microbial cells from new infections by closely related phages. Due to the protection from lysis and other infections, prophages can drive bacterial evolution by transfer of genetic information between multiple hosts and promote thereby an increased diversity [107,[111], [112], [113]]. The gene transfer can affect the capacity for biofilm formation, the abilities of hydrocarbon-degradation, antibiotic resistance or the virulence in a positive or a negative way [103,114].

To date, the natural occurrence of bacteriophages and their interactions with bacteria in natural oil reservoirs has not been studied in much detail. As far as we are aware, no viruses have been directly isolated from an oil reservoir. Only studies at a genomic level have revealed hints of the presence of viruses. Oil-water mixture samples from a production well of the water-flooded Chinese Qinghai oilfield were compared in taxonomic and functional compositions of the microbial communities in the oil and water phases by pyrosequencing and application of a GeoChip4.0 [1]. In that study, 38 of 40 detectable virus genes were found. Three of the detected genes showed significant differences in abundance between crude oil and water phase. Holin type 3 for bacterial lysis was more abundant in the oil phase, while the host recognition T2 type and the sliding clamp T4 for replication were both greater in the water phase. Because of the higher abundance of phage genes in the water phase, it was hypothesized that microbes are protected from phages by the oil phase [1]. Yet, holin type 3 abundance was higher in oil than in water phase, indicating phage-interactions and bacterial lysis directly within the oil phase [1]. As a paradigm of life, microorganisms can only live in a water phase. Thus, microbes in the oil phase are either physically partitioned into the oil during the production process, where they cannot live, or they are present in microdroplets of water dispersed in the oil. In the latter case, they would again be subject to viruses if these are present in the droplet. Another genomic study on Thermococcus sibiricus isolated from oil concluded that the oil environment is poorly invaded by bacteriophages because they only found a single CRISPR containing 24 repeat spacer units. However, it is known that other species of the order Thermococcales harbor multiple CRISPR loci carrying more repeat spacer units [115].

The studies presented have shown that viruses occur in oil reservoirs. However, questions concerning their ecological importance and the extent to which they shape and control microbial communities and processes remain to be elucidated. Studies on oil-contaminated waters such as spills and plumes and the correlated bioremediation processes propose a phage-driven microbial loop [105,116]. The authors proposed that phages ensure a persistent nutritional biomass turnover enabling bacterial hydrocarbon degradation [117,118]. We could not find any support of either ‘killing the winner’ or ‘piggyback-the-winner’ processes in natural oil reservoirs in the literature indicating an open field for ecological research.

Conclusion

Microbial oil degradation in deep subsurface oil reservoirs mainly takes place at the so-called oil-water transition zone (OWTZ) or oil-water interface. Even, if the OWTZ is a degradation hotspot, we propose that microbial oil degradation also takes place in small water-saturated parts of the rock containing actively living microbes or even in the thin water film of water wet reservoirs. Thus, biodegradation is distributed in a gradient through the entire oil field starting from the OWTZ and following the water content to the top of the reservoir. Therefore, the water inclusions should be considered as having a notable impact on overall oil degradation process in the deep subsurface.

Water samples from oil reservoirs obtained by pumping comprise a foamy mixture of oil, formation or injection water, and gas. We propose to examine both the water and the emulsified water within the oil phase of a sample to get a better picture of the entire community present in an oil reservoir. However, production water samples cannot provide information about the real distribution of the microorganisms in the deep biosphere but may provide insight into which organisms are involved in oil degradation.

Another important factor are syntrophic interactions between different microorganisms. Often, planktonic microorganisms do not have access to both electron donor and electron acceptor as prerequisites for microbial energy conservation, especially fermenting bacteria and methanogenic archaea. Therefore, we suggest that the methanogenic degradation of oil mainly takes place in mutualistic microbial consortia organized in biofilms where fermenting microbes transfer electrons either directly or indirectly to the methanogens. In laboratory experiments, many oil-degrading enrichment cultures and isolates build biofilms on the surfaces of alkane phases or PAH´s. Since known advantages of biofilms include protection against toxic compounds and desiccation and syntrophic electron transfer between fermenting organisms and methanogenic archaea, we propose that life in deep subsurface oil reservoirs is arranged predominantly in biofilms.

An important influence regarding degradation rates of crude oil is reservoir temperature, with most inhabited reservoirs ranging from mesophilic to thermophilic conditions. The largest microbial diversity occurs at moderate temperatures of up to 55 °C, where higher metabolic activity and increased abundance of genes involved in carbon cycling and the degradation of aromatic and other organic compounds occurs. Above ∼80 °C, oil reservoirs are considered to be sterile.

Oil reservoir temperature gives us an idea of microbial temperature preferences on a genus level but not always on an order level and so far no general predictions can be made about the phylum or class level. Some organisms may serve as indicators for in situ reservoir temperature determination. However, general predictions on microbial community composition based on temperature alone are not feasible.

Viruses could be another important factor in oil reservoirs. Regarding the ’killing the winner’ and the ‘piggyback-the-winner’ hypotheses, viruses could have an impact on shaping microbial communities and function, but no concrete evidence from oil reservoirs has been provided thus far.

Oil reservoirs provide an exceptional habitat for microorganisms, influenced by abiotic and biotic factors. Over the last decades, knowledge on the oil microbiome has grown but the function of the microorganisms described and the principles of the microbial oil degradation process still constitute open questions.

Funding sources

This work was supported by the European Research Council (ERC) [grant number 666952-EcOILogy] and the German Research Foundation (DFG) [grant number BR 5493/1-1].

Conflict of interest

The authors declare no conflict of interest.