In the era of Industry 4.0, with global climate change, increasing population and developing technology, the spread of heavy metal pollutants in aquatic areas is increasing. Bacterial resistance and metal accumulation capability are common phenomena that can be exploited for the bioremediation of the environment, hence these resistant bacteria may be potential candidates for biotechnological applications. Despite the risks caused by antibiotic-resistant bacteria, heavy metal-resistant bacteria can be used in detoxification processes to convert a toxic form to a harmless form of a substance by developing biotransformation mechanisms. Bioremediation studies have been carried out to identify candidate species (1–4).

In recent years, the increase in pollution by toxic compounds and heavy metals in marine areas makes it increasingly important to study the relationships between bacteria and toxic compounds. Studies related to the transformation of compounds into different forms via bacterial metabolic processes for the removal of toxic substances from the environment have gained importance. Detection of bacteria that are resistant to heavy metals in natural environments constitutes the first step to provide data for remediation studies.

Bacteria are some of the most important components in marine ecosystems. Since bacteria adapt to new conditions created by environmental variables around them, knowledge of bacteria provides data in terms of defining environmental factors, public health status and ecosystem function. Marine areas are exposed to domestic and industrial wastes depending on local technology levels and population. Many xenobiotic micro pollutants, antibiotic derivatives and metabolites reach the sea from human activity. This concerning issue is considered an important factor for global health with respect to the evolution and detection of antibiotic resistance in bacterial pathogens (5). Since the spread of antimicrobial resistance is not restricted by phylogenetic, geographic or ecological borders, studies describing regional status of bacterial resistance in natural areas are important.

Antibiotic resistance can spread rapidly among bacterial species (6). It is known that the occurrence of antibiotic-resistant bacteria in natural environments reduces the effectiveness of antibiotics in the treatment of infectious diseases. Due to the increasing global resistance of bacteria against antibiotics, humanity is constantly being forced to develop new antibiotic derivatives. This vicious circle is one of the most important problems of our age and poses a threat for the future. Thus, it is important to know the resistance levels of bacteria and to produce regional antibiotic resistance profiles in natural areas. Aquatic environments constitute a way to disseminate not only antibiotic-resistant bacteria but also the resistant genes in natural bacterial habitats (7). It has been well documented that the aquatic environment is a potential reservoir of antibiotic-resistant bacteria, furthermore the prevalence and persistence of antibiotic resistance in bacterial pathogens is a threat and a source of considerable concern to public health (8–15). It is known that environmental factors such as overpopulation, livestock farming, insufficient drainage and sanitation infrastructure may provide hotspots for environmental antibiotic-resistant bacteria transmission (16).

In aquatic environments, antibiotic-resistant bacteria can be accompanied by heavy metal-resistant bacteria that are often induced by the presence of metal caused by anthropogenic activities and environmental factors (16, 17). Heavy metals are introduced into the marine environment in different ways. Accumulation in sediment can affect aquatic life negatively for a long time. Bacteria that will take part in the transformation of heavy metal salts into harmless forms must be resistant to the heavy metals. Bacteria that cannot adapt to the changes metabolically will be eliminated and therefore various pollution inputs accumulated in the sediment will affect the composition of microbial diversity. Sediments containing harmful, inorganic or organic particles are relatively heterogeneous in terms of physical, chemical and biological properties and are an important source of heavy metal contamination (11). It has been reported that microplastics mediate the spread of metal- and antibiotic-resistant pathogens due to their ability to adsorb various pollutants (18, 19). Bacteria resistant to heavy metals in marine areas have developed various resistance mechanisms to counteract heavy metal stress. Only bacteria that can withstand the current heavy metal concentration can survive in these areas.

Heavy metals accumulate in biota via food chains and are transferred between organisms in marine environments. This cumulative process, named biomagnification, is higher in the sea than in terrestrial environments (15, 20) and this implies significant effects of heavy metal pollution in marine areas. On the other hand, heavy metal-resistant bacteria can play a role in detoxification by converting a toxic form into a harmless form through biotransformation mechanisms that develop in natural environments. These mechanisms include the formation and sequestration of heavy metals in complexes and the reduction of a metal to a less toxic species (21, 22).

Metal-resistant bacteria have developed very efficient and varying mechanisms for tolerating high levels of toxic metals and thus they carry an important potential for controlling heavy metal pollution (23). In many prokaryotes, it has been shown that the mechanism for resisting heavy metals develops over time. This process has been studied in species such as Escherichia coli and Staphylococcus aureus. It is reported that many different species of Pseudomonas, Bacillus, Enterobacter, Providencia and Chryseobacterium are efficient for reducing heavy metals (1–4).

It is known that the occurrence of bacteria resistant to antibiotics and heavy metal salts in the sea is related to the pollutants present in the environment. For the reasons highlighted above, it is important to determine the profile of antibiotic and heavy metal-resistant bacteria in marine environments. Marine areas which have different environmental inputs present novel media for bacterial studies.

For the present study, the Güllük Bay of the Aegean Sea, Turkey, was chosen since it is a dynamic area due to marine transportation, seasonal population growth depending on tourism, aquaculture, recreational and agricultural activities and terrestrial pollution inputs transported from rivers.

Probable faecal source analysis conducted in Güllük Bay showed that the primary source of the detected bacteriological pollution is anthropogenic (24). A significant part of domestic wastewater in the region collects in sealed septic tanks. It is possible for the wastewater to reach the sea by mixing the sedimentary septic tanks with groundwater. Chemical and biological studies (24–33) confirm that regional pollutants have reached Güllük Bay.

It is well known that sewage transported via domestic wastewater carries antibiotics to marine environments. This has an effect on metabolic capabilities of bacteria in marine environments. For example, β-lactam antibiotic derivatives used for human infection treatment may enter marine environments via domestic wastewater. Bacteria may obtain resistance via intercellular contact mostly using a conjugation mechanism (34). The existence of antibiotic-resistant bacteria is an indicator of domestic pollution. Furthermore, antibiotic-resistant bacteria may cause a vicious cycle. This problem has grown in recent years due to systematic use of antibiotics in animal husbandry and overuse of antibiotics (35, 36).

The frequencies of heavy metal-resistant bacteria and antibiotic-resistant bacteria were investigated in seawater and sediment samples collected from Güllük Bay in the period between May 2011 and February 2013.

2.1 Sampling Area

2.2 Sampling

2.3 Bacterial Isolation and Identification

2.4 Bacterial Resistance Against Antibiotics

2.5 Bacterial Resistance Against Heavy Metal Salts

3.1 Bacterial Resistance Against Antibiotics

3.2 Multiple Antibiotic Resistance Indexes

3.3 Bacterial Resistance Against Heavy Metals

Indicator bacteria levels reported in Güllük Bay and the presence of pathogenic bacteria (25, 26) support the relationship between the resistance data detected in the current study with bacteriological pollution levels. In the present study, bioindicator bacteria showing human-induced pollution input isolated from seawater had the highest frequency of resistance against nitrofurantoin (100%) and sulfonamide (95%). Sulfonamides were the first antibiotics developed for clinical use. Sulfonamides have been widely used to treat bacterial and protozoan infections in humans, domestic animals and fish since their introduction to clinical practice in 1935 (45–47). The results of higher resistance against sulfonamide in the present study were similar to the findings of sulfonamide resistance in another study (48). For example, there were significant increases in numbers of bacteria resistant to oxytetracycline, oxolinic acid and florfenicol in sediments from an aquaculture site compared with those from a non-aquaculture control site. Interestingly, in another study a similar number of antibiotic-resistant bacteria were isolated from aquaculture and non-aquaculture sites (49). Gram-negative bacteria (predominantly Plesiomonas shigelloides and Aeromonas hydrophila) were isolated from aquaculture ponds in the south-eastern USA and it was reported that antibiotic resistance to tetracycline, oxytetracycline, chloramphenicol, ampicillin and nitrofurantoin were higher in antibiotic-treated ponds compared to non-treated rivers (50). It was determined that bacteria isolated from Sopot Beach, Poland, were resistant to ampicillin (51). A high percentage of bacteria were reported as resistant to streptomycin (100%), cefazolin (89.8%), ampicillin (83.7%) and trimethoprim-sulfamethoxazole (69.4%), whereas a low percentage of bacteria were resistant to cefepime (12.3%) and meropenem (14.3%) in the aquaculture region of İskenderun Bay, Turkey (52).

In the current study, higher numbers of sulfonamide, rifampicin and ampicillin-resistant bacteria were recorded in the stations around aquaculture areas than other stations. Sphingomonas paucimobilis, Escherichia coli and Enterobacter cloacae isolated from both seawater and sediment at the stations around aquaculture areas had the highest levels of antibiotic resistance. The development of resistant pathogens in aquaculture environments is well documented (53, 54) and evidence of transfer of resistance encoding plasmids between aquaculture environments and humans has been presented recently (55). It has been reported that antibiotic‐resistant bacteria are present in a seafood ecosystem where antibiotics have never been used (56). This is interesting in terms of showing that aquaculture areas may be adversely affected by the presence of environmental antibiotic-resistant bacteria.

In the present study, a high percentage of the bacteria Sphingomonas paucimobilis were isolated, which was especially prevalent in Güllük Bay. The natural habitat of Sphingomonas has not been defined, but it is widely distributed in the natural environment especially in water and soil (57). The second most prevalent species were Escherichia coli and Enterobacter cloacae. Escherichia coli is an indicator of faecal contamination in aquatic environments. Enterobacter cloacae is the most frequent species associated with nosocomial infections along with Klebsiella pneumoniae that is a growing problem in human healthcare. The highest number of Bacillus cereus was isolated from the sediment underneath fish farms. A few Bacilli of marine origin have been reported to produce unusual metabolites different from those isolated from terrestrial bacteria (58). Due to the ubiquity and ability of the Bacillus species to survive under difficult circumstances, Bacillus strains are considered to be species of certain habitats (59, 60). In the current study, Bacillus pumilus, B. thuringiensis, B. mycoides and B. cereus were isolated from the sediment samples of the stations around fish farms.

The high frequency of resistance among bacterial isolates in the present study confirms the earlier reports regarding the role of antimicrobial use that plays a role in selecting antibiotic-resistant bacteria in water and aquatic sediments (46–52). Many previous studies have shown that the increases in antibiotic resistance in human medicine, agriculture and aquaculture are directly related to the amounts of antimicrobials used (61–65).

Infections caused by antibiotic-resistant bacteria are one of the most important public health concerns worldwide. Currently, MARs have been reported in a wide range of human pathogenic or opportunistic bacteria such as Vibrio sp. (66), Klebsiella pneumoniae (67), Salmonella sp. (68), Pseudomonas aeruginosa and also in pathogens (69, 70). Reservoirs of antibiotic resistance can interact between different ecological systems and potential transfer of resistant bacteria or resistant genes from animals to humans may occur through the food chain (70). In the current study, the MAR index of multiple antibiotic-resistant bacteria was found to be 2.6 times greater in the stations around fish farm areas (0.057) than the other stations (0.022).

Marine sediments offer more informative results than seawater about environmental pollution due to the accumulation of various pollutants at the bottom of the sea, therefore analysis of sediments is widely used in tests. The association of microorganisms with sediment particles is one of the primary factors in assessing microbial fate in aquatic systems. In this study, the bacteria isolated from sediment in all samples showed a higher resistance rate than bacteria isolated from seawater. Detection of higher antibiotic resistance in sediment bacteria than bacteria isolated from seawater showed that sediment bacteria were exposed to more antibiotics. Natural ecosystems containing high concentrations of heavy metals are also frequent. Heavy metal resistance genes are commonly found in environmental bacteria (71). The resistance to seven heavy metals has been reported in the order Cu > Mn > Ni > Zn > Pb > Cd > Fe for seawater bacteria isolated from the Golden Horn, Istanbul, Turkey (17). Heavy metal resistance in bacteria found in seawater from the Mediterranean has been reported as Cd > Cu > Cr = Pb > Mn; in Karataş, Turkey Cd > Cu > Cr = Mn > Pb; and İskenderun Bay, Cu > Cd > Mn > Cr > Pb (72).

In the present study, resistance to five different heavy metals (Zn²⁺, Pb²⁺, Cu²⁺, Cr³⁺ and Fe³⁺) were investigated for all isolates. Trends in heavy metal resistance vary depending on the sample sites: Güllük Bay, fish farm water column: Cu > Zn > Pb > Cr > Fe; sediment: Cr > Cu> Zn > Fe > Pb. Frequency of bacteria resistance to heavy metals shows the direct effects of metal pollution. Neisseria animaloris, Aeromonas caviae and Bacillus cereus isolated from sediment samples were the most tolerant of all the heavy metal salts. Chryseobacterium indologenes displayed the highest degree of sensitivity to all metal salts while Lactococcus garvieae showed the highest degree of sensitivity to Zn²⁺, Pb²⁺, Cu²⁺ and Fe³⁺. Kocuria kristinae, Escherichia coli and Acinetobacter lwoffii, which were isolated from the seawater underneath the fish farm, displayed similar sensitivities to all tested heavy metal salts. Resistances to heavy metals for Aeromonas and Pseudomonas isolates were similar to those from İskenderun Bay, with cadmium, 35.0% and 56.5%; copper, 98.3% and 75.4%; chromium, 38.3% and 31.9%; lead, 1.7% and 7.2%; manganese, 43.3% and 44.9%; and zinc 35.0% and 41.3%, respectively (72).

Both Gram-positive and Gram-negative bacteria can resist heavy metals (73). Resistance to toxic metals in bacteria probably reflects the level of environmental contamination with these substances and it may be related to the concentration of bacteria (74). The present project found heavy metal pollution in Güllük Bay sediment samples at all stations. In the sediment samples, the heavy metal contents were reported at varying rates: between 1 μg g⁻¹ and 209 μg g⁻¹ for lead; 10 μg g⁻¹ and 259 μg g⁻¹ for zinc; 1 μg g⁻¹ and 59 μg g⁻¹ for copper; 0.1 μg g⁻¹ and 46 μg g⁻¹ for chromium; <0.01 μg g⁻¹ and 2.8 μg g⁻¹ for cadmium; <0.01 μg g⁻¹ and 0.4 μg g⁻¹ for arsenic; and 0.6% and 5.9% for aluminium, respectively. The region was defined according to cadmium, lead and zinc levels as moderately polluted. Recorded high metal values were evaluated as an indicator of domestic and industrial inputs, carried via Sarıçay Creek, port operations and tourism activities within Güllük Bay (75). In the current study, the high frequencies of heavy metal-resistant bacteria detected in the sediment samples support this data. Bacterial heavy metal resistance detected in the study may depend on many factors. A possible explanation for differences in heavy metal resistance is the proximity of Güllük Bay to iron-steel factories. Additionally, Güllük Harbour is a serious pollution source. It was reported that 2862‐unit ships carried 4.8 million tonnes of ballast water to Güllük Harbour during 2007–2012 (37). Another potential source of increased resistance may be the discharge of thermal power plants located 107 km, 46 km and 39 km away from Güllük Bay. The effects of thermal power plant discharge on the accumulation of heavy metals have been reported in other studies (29, 75).

The association between antibiotic resistance and resistance to heavy metals is quite common in the same organism. The increasing numbers of antibiotic and heavy metal-resistant bacteria could be a result of gene transfer activities demonstrating that industrial pollution most likely selects for antibiotic resistance and vice versa (58). In this study, similarly, the most antibiotic-resistant bacteria such as Sphingomonas paucimobilis, Escherichia coli and Enterobacter cloacae were also resistant to heavy metals. Metal‐resistant isolates from Güllük Bay also showed high resistance to sulfonamide, rifampicin and ampicillin. Bacteria from different sources such as humans, animals and soil can transfer or exchange their resistance genes. At the same time, water contaminated with antibiotics, disinfectants, pesticides and heavy metals might encourage selection and result in antibiotic and heavy metal resistance. Marine environmental conditions are extremely dynamic compared to the terrestrial environment, allowing bacteria to bring resistance mechanisms they have developed together while being adapted to the varying conditions. This makes the isolation of various bacteria useful to assess environmental pollution and provides a pathway to possible solutions to remove pollution from marine environments. For bacteria to take part in the transformation of any heavy metal salt into a harmless form, those bacteria must firstly be resistant to the heavy metal; thus the data related to frequency of metal resistant bacteria can provide knowledge on the continual accumulation or transformation of heavy metals in the marine environment.

The findings of the current study provide data regarding the distribution of heavy metal- and antibiotic-resistant bacteria in seawater and sediment samples of Güllük Bay, Aegean Sea, Turkey. As a result, preliminary data on candidate bacteria will offer opportunities for further studies on the elimination of heavy metal contamination by the detection of heavy metal-resistant bacteria.

Analyses of the presence of antibiotic resistance in bacteria provide knowledge on pollution sources such as septic systems on regional ecosystems. Since antibiotic-resistant bacteria can affect pathogen virulence, these pollution sources can induce pathogens and can create health risks for both humans and the ecosystem. In the present study, bacteria resistant to antibiotics and heavy metals in seawater and sediment were investigated. The bacterial information obtained provides essential data for identifying the regional distribution of resistant bacteria. Levels of resistance against heavy metals and antibiotics in bacteria isolated from seawater and sediments of the Aegean Sea were quantified. Bacteria isolated from Güllük Bay sediment were resistant to all antibiotics tested and exhibited higher resistance than those isolated from seawater. The frequency of antibiotic-resistant bacteria was higher around fish farms and near the exit of Sarıçay Creek. The widespread resistances of indicator bacteria to antibiotics suggest the presence of anthropogenic influences due to domestic waste and maritime transport.

In order for bacteria to take part in the transformation of heavy metal salts into harmless forms, they must initially be resistant to heavy metals. The frequency of resistance thus provides information regarding the continual accumulation or transformation of heavy metal salts in the marine environment. The findings of the present research have shown the existing contamination status of Güllük Bay via heavy metal and antibiotic resistance tests. The study region is under pressure of pollution as stated in previous research (25, 26, 75) and the bacterial resistance data of the current study showed that there is a prevalence of resistant bacteria in the region that may be due to indirect effects of environmental dynamics and pollution.

In this study, the presence of higher levels of resistant bacteria in sediment compared to seawater may indicate the presence of microplastics in the sediment as well as the probability that the sediment is a suitable medium for accumulation of metals and antibiotics. Further studies on this subject will provide detailed data on the spread of antibiotic- and metal-resistant bacteria in marine sediments.

The present study showed bacterial responses to environmental stress and influences in terms of antibiotic and heavy metal resistance both in sediment and seawater samples at Güllük Bay, Turkey. These findings highlight the necessity of holistic assessments with a ‘one health’ approach and the need to control bacteria entering marine areas due to human activities, considering the contributions of resistant bacteria to global distribution. The data may also provide a useful resource to help identify strains of bacteria for environmental remediation applications.

Three soak liquor samples were collected from Istanbul Leather Organized Industrial Zone, Tuzla, Istanbul, Turkey. These samples were immediately placed into sterile sample bags and carried on ice during transportation. Direct and serial dilutions were spread onto nutrient agar plates. The morphologically different colony was picked up to obtain the pure culture of the isolate and was numbered as Isolate 1.

Gram staining, catalase, oxidase, lipase and protease activities were examined. Protease activity of Isolate 1 was examined on gelatin agar medium containing 2% gelatin (w/v). The agar plates were flooded with Frazier solution following 24 h incubation. Clear zones around the colonies were evaluated as positive for protease activity. Lipase activity was tested on Tween^® 80 agar medium containing 1% (w/v) Tween^® 80. After incubation, opaque zones around the colonies were accepted as evidence of lipase activity (29, 30).

Genomic DNA of Isolate 1 which was determined to have protease and lipase activities were extracted by phenol/chloroform extraction and ethanol precipitation. DNA isolation was confirmed by agarose gel electrophoresis. DNA samples were stored at −20°C until use. The 16S rRNA gene was amplified by polymerase chain reaction (PCR) with the universal bacterial primers 27F (5-AGAGTTTGATCMTGGCTCAG) and 1492R (5-TACCTTGTTACGACTT). Negative control was included in PCR amplifications. PCR amplification was carried out by an initial denaturation at 95°C for 4 min, followed by 30 cycles at 95°C for 1 min, 57°C for 1 min and 73°C for 1 min. The reactions were finished by a final extension at 73°C for 7 min. The PCR products were also monitored by agarose gel electrophoresis. These products were purified by GeneJET^TM Gel Extraction Kit (Thermo Scientific^TM, Thermo Fisher Scientific, USA). These purified samples were analysed by Medsantek Ltd Co, Istanbul, Turkey. The 16S rRNA sequence contigs were generated by the software ChromasPro version 2.1.8 (Technelysium Pty Ltd, Australia). Then, consensus sequences were exported in FASTA format for each sample for data analysis. These sequences were compared with sequences in the National Center for Biotechnology Information (NCBI) using the Basic Local Alignment Search Tool (BLAST^®) search program.

The lichen samples belonging to H. tubulosa, H. physodes, E. divaricata, P. furfuracea, P. sulcata and Usnea sp. were collected from fir trees of Kastamonu province in the north-west of Turkey. They were identified through classical taxonomical methods by microscopic examination.

H. tubulosa, H. physodes, E. divaricata, P. furfuracea, P. sulcata and Usnea sp: Turkey, Kastamonu province, Kapaklı Village, 41.24492, 34.18330, G. Çobanoğlu.

The experiment steps included washing, drying in air, weighing, pulverising by liquid nitrogen, adding acetone (ACS, ISO, Reag. Ph. Eur.), keeping in a dark place for 24 h followed by filtration through filter paper. Then, the evaporation of acetone in a rotary evaporator was performed and crude lichen acetone extracts were obtained (27).

The test isolate was grown on Tryptic soy agar media at 37°C for 24 h. The tests were performed in 96-well CELLSTAR^®, F-bottom microplates with lid (Greiner Bio-One GmbH, Austria). Tryptic soy broth was added to each well and nine-fold serial dilutions of the acetone extracts of H. tubulosa, H. physodes, E. divaricata, P. furfuracea, P. sulcata and Usnea sp. were made. Final concentrations of all lichen extracts were 240 μg ml⁻¹, 120 μg ml⁻¹, 60 μg ml⁻¹, 30 μg ml⁻¹, 15 μg ml⁻¹, 7.5 μg ml⁻¹, 3.75 μg ml⁻¹, 1.9 μg ml⁻¹ and 0.9 μg ml⁻¹. Overnight culture of the isolate was added to obtain a total volume of 100 μl with an optical density (OD) 600 nm of 0.01. The experiments included untreated and blank controls. The tests were performed in three replicates. Bacterial growth ratios at an OD 600 nm were measured using Cytation^TM 3 Multi-Mode microplate reader (BioTek Instruments Inc, USA).

Two deteriorated salted sheepskins containing red discolorations were collected from two tanneries in the Istanbul Leather Organized Industrial Zone (40°52′39.7″N,29°20′25.3″E) in Tuzla, Turkey. The samples were immediately placed into sterile sample bags and transported on ice to the laboratory. Then, 20 g of the salt-pack cured sheepskin samples were weighed and separately soaked in flasks containing 180 ml 30% NaCl (Merck KGaA, Germany) solution. The flasks were placed into a shaking incubator at 90 rpm, 24°C for 3 h. The suspension of the skin was diluted with sterile physiological saline water (30% NaCl). An aliquot of 100 μl each of direct and serial skin suspension dilutions was spread onto the surface of modified Brown agar media containing (per litre): 1 g CaCl₂·H₂O, 2 g KCl, 20 g MgSO₄·7H₂O, 3 g trisodium citrate, 250 g NaCl, 5 g yeast extract, 20 g agar, pH 7 (5, 27). The plates were incubated at 39°C for 10 days. Following incubation, red pigmented colonies on the agar media were selected and restreaked several times to obtain pure cultures. A total of 22 isolates were obtained from the sheepskins and then, these strains were examined the proteolytic and lipolytic activities. Proteolytic activity of each strain was detected on gelatin agar medium containing 2% gelatin. After incubation, clear zones around the colonies on the gelatin agar medium indicated protease production (5, 10). Lipolytic activity of each strain was screened on Tween^® 80 agar medium containing 1% Tween^® 80. After growth was obtained, opaque zones around the colonies were interpreted as positive lipase activity (5). In the present study three red pigmented proteolytic and lipolytic strains (AT1, 22T6 and 7T3) were obtained from two salted sheepskins and these strains were used in the present study.

Exponentially growing pure cultures of three strains designated as AT1, 22T6 and 7T3 were used in all experiments. First, the strains’ salt requirement and salt tolerance were examined on Brown agar plates containing different salt concentrations (0%, 0.5%, 3%, 5%, 7.5%, 10%, 12.5%, 15%, 20%, 25% and 30%) (27). After detection of optimum salt concentration for each strain, pH and temperature ranges for growth of each strain (AT1, 22T6, 7T3) were respectively examined at Brown agar plates with different pH values (pH 4, pH 5, pH 6, pH 7, pH 7.5, pH 8, pH 9, pH 10, pH 11 and pH 12) and different temperatures (4°C, 10°C, 15°C, 24°C, 28°C, 35°C, 37°C, 39°C, 45°C, 50°C, 55°C, 60°C) according to the methods described in Proposed Minimal Standards for Description of New Taxa in the Order Halobacteriales (28). Based on the pH, and temperature range of each test strain, the optimal pH and growth temperature of each test strain were determined.

Pigmentation, size, margin, elevation and opacity of colonies of the strains grown on Brown agar media were examined under optimal growth conditions (28). Cell morphology, cell length, cell width and motility of each strain were examined using both light microscopy and electron microscopy. Microscopic observation of each strain was made by using freshly prepared wet mount (28). For SEM observations, 20 ml of each test strain were separately passed through 0.2 μm pore size cellulose nitrate membrane filter placed in the stainless steel funnel via vacuum pump (Sartorius AG, Germany). The archaeal cells of each strain trapped on the membrane filters were observed under SEM (Quanta^TM 450 FEG (FEI, USA)). Gram staining was performed with acetic acid-fixed slides (28–30). Catalase and oxidase activities, indole production, methyl red test, H₂S and NH₃ productions of each strain were investigated according to the procedures described previously (4, 28, 31). Furthermore, each strain’s caseinase activity was determined on the agar medium containing 2% skim milk. After incubation, clear zones around the colonies were evidence of positive caseinase activity (4). Urease production was investigated on Christensen urea agar medium. The tubes were examined for pink or red colour change in the medium after seven days of incubation (28, 31). β-galactosidase activity was screened in test tubes containing ortho-nitrophenyl-β-galactoside (ONPG) discs and 1 ml of sterile saline water (30% NaCl). The yellow colour formation in the test tube was accepted as positive β-galactosidase activity (5, 31). Amino acid utilisation of each strain was examined in the test medium containing 1% amino acid, 0.5% beef extract, 0.5% peptone, 0.05% dextrose, 0.0005% cresol red, 0.001% bromocresol purple, 0.0005% pyridoxal and saline water (30% NaCl). Purple colour formation in the test tube containing archaeal culture was accepted as a positive test after 10 days incubation period at 39°C (31).

Chromosomal DNA was isolated by QIAamp DNA Mini Kit (Qiagen, Germany) and purified by QIAquick PCR Purification Kit (Qiagen, Germany) according to the manufacturer’s directions. The 16S rRNA genes of the strains were amplified by polymerase chain reaction (PCR) using forward primer 21F and reverse primer 1492R (32). The 16S rRNA gene sequences of three strains (AT1, 22T6 and 7T3) were determined by IONTEK Laboratory (Turkey). The sequences of these strains were analysed using ChromasPro v.2.1.8 software (Technelysium, Australia) and then compared with the sequence on the EZBioCloud Database (ChunLab, South Korea) (33).

2.4.1 Preparation of Strains and Cultures Used in Brine Curing Processes

2.4.2 Preparation of Sheepskin Samples for Brine Curing Treatments

To determine total counts of extremely halophilic archaea in the curing solutions before the curing processes, 100 μl of the test medium was removed from the each curing solution and diluted to 10⁻²–10⁻⁴ using sterile 30% NaCl solution. The diluted archaeal suspensions were spread over the Brown agar media. In addition, subsequent to each brine curing process detailed above (T1–T5), the suspensions of cured sheepskin samples were prepared at intervals of 5 days, 16 days, 28 days and 47 days of storage. 2 g of each skin sample were put into a flask containing 18 ml sterile 30% NaCl solution and incubated for 1 h at 24°C and 100 rpm. Direct and serial dilutions of the suspensions were spread onto the surface of Brown agar media. All inoculated Brown media were incubated at 39°C for 10 days and the colonies grown on the test media were counted.

After curing processes, 5 g of the sheepskins were cut and placed into flasks containing 100 ml of sterile distilled water. The flasks were placed in a shaking incubator for 1 h at 100 rpm and then pH was measured with a pH meter. Hairs and dirt on the samples were removed to properly determine the samples’ moisture content. 3 g of the samples were placed into an oven at 102°C for 6 h. The dried samples were weighed, returned to the oven for 1 h, and then were weighed again. The drying procedure was repeated until the first dry weight was equal to the second dry weight. The samples were put into a desiccator for 30 min to cool. Next, we calculated the skins’ moisture contents (20, 21). The dry sheepskins samples were placed in ceramic crucibles and ashed in a muffle furnace at 600°C for 8 h. After cooling, the samples were weighed to determine ash content. Moisture content, ash content and salt saturations of skin samples were calculated according to the aforementioned methods (30, 36). The pH value, ash content, moisture content and salt saturation of all cured sheepskin samples were examined at different storage periods.

All cured sheepskin samples were examined organoleptically (hair slip, deterioration of skins, bad odour, sticky appearance, red heat, hole formation) during different storage periods.

After a 47-day storage period, the sheepskin samples were prepared for SEM observation. The samples were fixed in 4% glutaraldehyde solution prepared in 0.1 M phosphate buffer (pH 7.2) for 30 min. The samples were washed three times with 0.1 M phosphate buffer for 10 min and were treated with 1% OsO₄ prepared in 0.1 M phosphate buffer at room temperature for 1 h. The samples were washed two times in sterile distilled water for 10 min. Then, the water in the sheepskins was gradually removed by 35%, 50%, 75%, 95% and absolute ethanol. The mixtures of ethanol-hexamethyldisilazane (ethanol-HMDS) [1:1 (v/v)] (1 × 30 min), ethanol-HMDS [1:2 (v/v)] (1 × 30 min) and HMDS (2 × 30 min) were used for air drying process. After drying, HMDS was poured from petri dishes and the samples were placed in a desiccator for 12 h. Later, the sheepskin samples were examined under SEM (Quanta^TM 450 FEG) using sample stub with double-sided sticky tape (37).

A total of 22 red coloured strains were isolated from two deteriorated salted sheepskin samples obtained from two tanneries in the Istanbul Leather Organized Industrial Zone in Tuzla, Turkey. While nine, seven and three strains respectively produced protease, lipase, both protease and lipase, three strains did not produce either lipase or protease enzymes. The red coloured three extremely halophilic strains producing both protease and lipase enzymes were selected and used as test strains (AT1, 22T6 and 7T3) in the present study.

Strains AT1, 22T6 and 7T3 grew at 15–30% NaCl, 15–30% NaCl, 20–30% NaCl concentrations, respectively. Optimum salt concentrations of strains AT1, 22T6 and 7T3 were determined as 25% NaCl. Hence, these strains were accepted as extremely halophilic archaea. The pH and temperature ranges for growth of strains AT1, 22T6 and 7T3 were respectively found as pH 6–11 and 20–50°C, pH 6–11 and 15–55°C, pH 5–11 and 15–55°C. All extremely halophilic archaeal strains optimally grew at 39°C and pH 7. The colony pigmentation, size, margin, elevation and opacity of strains AT1, 22T6, 7T3 were respectively observed as: red, 0.6– 2 mm, entire, convex, translucent; red, 1–2 mm, entire, convex, translucent; red, 0.8–1.9 mm, entire, convex, translucent. The cells of strains AT1 (Figure 2(a)) and 7T3 (Figure 2(c)) were non-motile, extremely pleomorphic (triangle, square, irregular disk, short rod). The cells of strains AT1 and 7T3 were approximately 0.4–1.3 μm × 0.4–2.0 μm and 0.3–0.7 μm × 0.3–4 μm, respectively. The cells of strain 22T6 (Figure 2(b)) were motile, pleomorphic rods, approximately 0.5–1.2 μm × 3.2–6.6 μm. All strains were Gram-negative (Table II). While all strains showed positive catalase, oxidase, protease, lipase activities, indole production, the methyl red, caseinase, urease and β-galactosidase reactions of all strains were negative. The strains did not produce H₂S and NH₃ (Table II).

Fig. 2

Our experimental results showed that Haloarcula salaria (AT1), Halobacterium salinarum (22T6), Haloarcula tradensis (7T3) strains have protease activities which can breakdown proteins in corium of sheepskin causing loss of skin substance. When the protein structure of salted skins is broken down by proteolytic extremely halophilic archaea, these microorganisms can utilise some amino acids as a source of carbon, nitrogen and energy. Haloarcula salaria AT1 and Halobacterium salinarum 22T6 utilised most of the amino acids examined. While Haloarcula salaria AT1, Halobacterium salinarum 22T6 utilised 17 amino acids, Haloarcula tradensis 7T3 used only three amino acids (Table III). In another study, the liquid test media containing calfskin samples, 30% NaCl and proteolytic red and pink strains of the extremely halophilic archaea were separately prepared to show disintegration of the skin proteins. After an incubation period, decomposition of the skin samples in the media was detected by visual observation. While contents of asparagine, threonine, serine, glutamine, proline, glycine, alanine, valine, isoleucine, leucine, phenylalanine, lysine and arginine in the test tubes were detected at high levels, contents of methionine, tyrosine and histidine were low (10).

Phenotypic features of extremely halophilic AT1, 7T3 and 22T6 strains detected in this study were fairly similar to phenotypic features of Haloarcula salaria, Haloarcula tradensis and Halobacterium salinarum isolated by other researchers (15, 38, 39).

The phylogenetic analysis revealed that three strains shared highly similar identities with their closest phylogenetic relatives. Strains AT1, 22T6, 7T3 were respectively assigned to Haloarcula salaria (98.36%-1344 base pairs), Halobacterium salinarum (99.78%-1345 base pairs), Haloarcula tradensis (98.37%-1355 base pairs). The gene sequence data of the strains AT1, 22T6, 7T3 were respectively deposited in GenBank^® (National Center for Biotechnology Information, USA) under accession numbers as MN585896, MN585803, MN585804.

In our previous study, extremely halophilic archaeal strains were isolated from Tuz Lake and its salterns (5). In Turkish leather industry, curing salt is mostly obtained from Tuz Lake and its salterns. Hence, we suspect that contaminations of our sheepskin samples with Haloarcula salaria AT1, Halobacterium salinarum 22T6 and Haloarcula tradensis 7T3 were due to the curing salt obtained from Tuz Lake and its salterns.

In the study carried out with 25 salted sheepskin samples (Australia, Bulgaria, Dubai, Greece, Israel, Kuwait, South Africa, Turkey, USA) and 25 salted goat skin samples (Australia, Turkey, Bulgaria, Israel, South Africa, Russia, China, France), proteolytic extremely halophilic archaea and lipolytic extremely halophilic archaea were detected as 10²–10⁵ CFU g⁻¹; 10²–10⁶ CFU g⁻¹ and 10²–10⁶ CFU g⁻¹; 10²–10⁶ CFU g⁻¹ on salted sheepskins and goat skins, respectively (40). The highest number of proteolytic and lipolytic extremely halophilic archaea on the salted skins was found as 10⁶ CFU g⁻¹ (40). Therefore, the archaeal cell numbers of test strains in the brine curing solutions were adjusted to 10⁶ CFU ml⁻¹. Before the curing processes of sheepskins, while the archaeal cell numbers in the brine solutions of Treatments 1, 3 and 4 were detected as 2.1 × 10⁶ CFU ml⁻¹, the archaeal cell numbers in the brine solution of Treatment 2 was detected as 2.2 × 10⁶ CFU ml⁻¹.

The archaeal cell numbers in the mixed culture was detected as 2.1 × 10⁶ CFU ml⁻¹ in the electrolysis cell before 1.5 A DC application. While the archaeal cell numbers in the mixed culture were reduced from 2.1 × 10⁶ CFU ml⁻¹ to 3.2 × 10⁵ CFU ml⁻¹ after 1 min of DC treatment, the cell numbers of 1.24 × 10² CFU ml⁻¹ was detected after 4 min of DC treatment. All archaeal cells in the mixed culture were completely killed in 7 min of DC treatment. In the present study, log₁₀ value of the mixed culture of extremely halophilic archaea in the brine solution before the DC treatment was 6.32. After 1 min, 4 min and 7 min of 1.5 A DC treatment; 0.82, 4.23 and 6.32 log₁₀ reduction values (CFU ml⁻¹) of the mixed culture in the brine were detected, respectively.

Temperature and pH of the electrolysis cell were respectively measured as 31°C and pH 6 prior to the electric current treatment. After treating the brine solution with the electric current, the temperature of the brine was adjusted to 24°C for using in curing process of sheepskin in the Treatment 5. While the temperature and pH of the test medium respectively increased from 31°C to 41°C and from pH 6 to pH 8.5 during the electric current treatment, voltage values slightly decreased from 4.7 V to 4.3 V.

We also demonstrated the inactivation of extremely halophilic strains via DC and alternating electric current (AC) in our previous studies (35, 41, 42). A 0.5 A DC was applied for 30 min to several strains of extremely halophilic archaea (10⁷ CFU ml⁻¹) isolated from Tuz Lake, Kaldırım and Kayacık salterns (35). While the mixed culture of extremely halophilic archaea was exterminated in 10 min, protease producing extremely halophilic archaea were killed in 5 min. However, lipase or lipase and protease producing extremely halophilic archaea were exterminated in 20 min (35). In another experiment, lipase and protease producing extremely halophilic strains (10⁵–10⁶ CFU ml⁻¹), separately grown in liquid Brown media, were inactivated by a 10 min treatment with 0.5 A DC (41). It was also detected that 1 min of 2 A AC treatment was enough to kill extremely halophilic archaea found in brine solution (10²–10⁴ CFU ml⁻¹). When 2 A AC was applied to lipolytic extremely halophilic archaea, proteolytic extremely halophilic archaea, both proteolytic and lipolytic extremely halophilic archaea, and a mixed culture of these strains (10⁶ CFU ml⁻¹), all test microorganisms found in 25% NaCl solution were exterminated in 5 min (42).

After the curing processes of sheepskins, we did not detect any extremely halophilic archaea on the sheepskin sample cured with the sterile brine solution (Control) and the sheepskin sample cured with the brine solution containing electrically inactivated mixed culture (T5) during the all storage periods.

While extremely halophilic archaeal numbers on both skin samples cured with each strain and the skin sample cured with mixed cultures of the strains slowly increased from 10⁶ CFU ml⁻¹ to 10⁷ CFU during five days and 16 days storage periods, the numbers of these strains slowly decreased 28 days and 47 days storage periods due to attachment of these cells to sheepskins (Table IV).

After the curing processes of skins, pH values of the sheepskin samples were measured as pH 7.35 for Control; pH 6.89 for T1; pH 7.09 for T2; pH 7.05 for T3; pH 7.16 for T4; pH 8.05 for T5. While salt saturation values of all cured sheepskins were higher than 100% during all storage periods, pH, ash content and moisture content values changed during different storage periods. pH, ash content and moisture content values of the cured skins were detected between pH 6.52–7.81, 20–44%, 24–57%, respectively (Table IV).

Moisture, minimum and maximum ash contents, salt saturation values of adequately cured salted hides were suggested as 40–48%, 14–48%, higher than 85%, respectively (36). Due to detection of high moisture content in all samples (between 50–57%) after five days storage, sterile salt was added to all sheepskins to reduce their moisture contents according to curing procedure described in the previous study (43). While all skin samples reached the suggested moisture content values (39–46%) after 28 days, the suggested saturation values were detected after five days. The samples’ lowest moisture content values were detected after 47 days. Ash contents of all skins (20–44%) were close to suggested values (36). While the skins’ pH values changed during storage periods, all values were found sufficient to support the growth of extremely halophilic strains (Table IV). The pH, moisture content, ash content and salt saturation values detected in this study were also consistent with pH range (pH 6.53–8.01), moisture content (32–68%), ash content (12–30%) and salt saturation (58–100%) values of 25 salted sheepskin samples determined in the previous experiment (40).

While hair slip and bad odour were detected on the sheepskin samples cured with each strain and the mixed culture after five days at 33°C, sticky appearance and red heat were observed on the cured sheepskin samples after 16 days (T1–T4, Figure 3). In addition to the aforementioned organoleptic properties, hole formations were observed on these sheepskin samples after 28 days. However, we did not detect any organoleptic properties on sheepskin samples cured with sterile brine and the brine treated with 1.5 A DC (Control and T5, Figure 3).

Fig. 3

In another study, the commercially cured hides stored one year in the USA were also examined for proteolytic activity of extremely halophilic archaea. Experimental results of that study showed that the flesh side of hides containing extremely halophilic archaea had pink discolorations called red heat. When these hides were incubated at 35°C–40°C, bad odour, hair slip and severe grain damage were detected. Damaged grain surfaces were observed on leather made from these hides (10). In another experiment researchers emphasised that temperatures of the brines and hides should be maintained below 20°C to prevent growth of extremely halophilic archaea (44).

Figure 4 shows extremely halophilic archaeal cells of the mixed culture on 0.2 μm pore-size cellulose nitrate membrane filter in pleomorphic shapes such as triangle, square, irregular disk and rod. As seen in the SEM micrograph, 1.5 A DC treatment significantly debilitated structural integrity of the cells in the mixed culture trapped on the filter (Figure 5). The SEM images clearly showed that electric current application damaged cell structures of each strain in the mixed culture (Figure 5). As seen in Figure 6, the sterile brine curing process protected the sheepskin against microbial damage during 47 days of storage.

Fig. 4

Fig. 5

Fig. 6

Attachment of Haloarcula salaria AT1, Halobacterium salinarum 22T6 and Haloarcula tradensis 7T3 to corium fibres and the consequent destructive effects on sheepskins are seen in Figures 7–10. Haloarcula salaria AT1, Halobacterium salinarum 22T6 and the mixed culture of the strains caused fibres in the corium to split and weaken (Figures 7, 8 and 10). In contrast with the skin samples treated with Haloarcula salaria AT1, Halobacterium salinarum 22T6, skin sample treated with Haloarcula tradensis 7T3 had compact appearance, although the shredding of the fibres was still present in corium (Figure 9). That damage was due to the proteolytic activities of these microorganisms.

Fig. 7

Fig. 8

Fig. 9

Fig. 10

Figure 11 clearly shows that the curing process of sheepskin with the brine containing mixed culture treated with DC prevented extremely halophilic archaea from contaminating the sheepskin and furthermore protected the skin very well against microbial damage during a long storage period.

Fig. 11

The present study proved that organoleptic changes detected in the sheepskins were closely related to proteolytic and lipolytic activities of extremely halophilic archaeal strains on the skin. Electron micrographs also showed that each test isolate and a mixed culture of extremely halophilic strains destroyed the skins’ collagen fibres. We did not detect any difference when assessing the efficacy of sterile brine and electrically treated brine curing processes of sheepskin samples throughout 47 days. We did not observe any damage to the compactness of sheepskin structure cured with both the sterile brine and electrically treated brine containing the mixed culture. Both methods were found very effective for preventing archaeal growth and damage on the brine cured sheepskins.

Our results were consistent with those of other experimental studies on the extremely halophilic strains and culture collection strains of extremely halophilic archaea (25, 26). In our previous experiment, SEM images showed that hides cured with proteolytic extremely halophilic archaeal strains had red heat and severe grain damage after 49 days of storage at 41°C (25). In another study, the cured hides with extremely halophilic Haloferax gibbonsii (ATCC^® 33959^TM) exhibited hair loss, thinner and flaccid structure; these consequences of deterioration and loss of hide substance. The open fibre structure was also detected in the corium of the hide inoculated with Haloferax gibbonsii (27). The SEM images showed that the fibre structures of hide were broken down into the smaller fibres after 43 days (27).

Parsık cave is located in Izmit-Aksığın village (Global Positioning System (GPS) coordinates 40° 37′ 50.1060″N, 29° 57′ 56.5056″E), in the north-west of Turkey. It is a horizontal cave with a length of 778 m and a depth of 166 m. There is an intense water inlet in Parsık cave throughout four seasons. Samples were taken from water, soil and surface formations (‘moonmilk’) (Figure 1). The selected sampling zones are the sole area away from the entrance area, trip and running water pathway. Although Parsık cave is not a show cave, it is open to cavers and researchers.

Fig. 1

Surface formation samples were collected by sterile swabs under aseptic conditions and cultivated on starch casein agar (SCA), inorganic salt-starch agar (ISP4), soil extract agar (SEA) and Actinomycetes isolation agar (AIA-G) in duplicate for each region. Once the plates reached the laboratory, they were incubated aerobically for a period of 5–30 days at 20°C (21). All water and soil samples were taken in sterile sample containers.

Humidity and temperature values of the sampling areas were measured by a portable temperature/humidity meter. In addition, the temperature, conductivity, amount of dissolved substances and pH values of the sampled water sources were measured during sampling and recorded by a HQ40D digital two channel multimeter (Hach Lange GmbH, Germany).

The redox dye 5-cyano-2,3-ditolyl-tetrazolium chloride (CTC) was used together with the DNA-binding fluorescent dye 4’,6-diamidino-2-phenylindole (DAPI) to determine the total number of bacterial cells and the viable count of bacteria which actively respire. The concept is to distinguish between the metabolically active cells and the dead cells present in each of the water and soil samples. The experimentation procedure is the same as previously described by Güngör and Yurudu (22).

1 l of water samples were condensed by using polyamide filters of 0.22 μm pore size. Filters were re-suspended in 20 ml of sterile physiological saline water. 1 g of the soil samples was homogenised in 9 ml of sterile physiological saline water. All samples were cultivated using the 10-fold serial dilution method. Diluted samples were cultured on tap water agar (TWA) and Reasoner’s 2A agar (R2A) for enumeration and isolation of bacteria from water and soil samples. In addition, bacterial isolation from soil samples was on SCA, ISP4, AIA-G, SEA and 1/2 tryptic soy agar (TSA) media, and that from water samples was on 1/2 TSA only.

Plates were incubated aerobically for a period of 5–30 days at 20°C (21). At the end of incubation, plates which contained between 30 and 300 colonies were considered for both soil and water samples. Colonies which appeared different were selected for identification, then stored at −86°C for subsequent uses.

Cave isolates were identified through biochemical tests performed in the VITEK^® 2 system (bioMérieux SA, France). One of the three formats of this system is the VITEK^® 2 Compact 30 which focuses mainly on the industrial microbiology-testing environment. Based on this industrial software, three reagent cards of VITEK^® 2 Compact 30, named Gram-negative fermenting and non-fermenting bacilli (GN), Gram-positive cocci and non-spore-forming bacilli (GP) and Gram-positive spore-forming bacilli (BCL), were used to characterise the isolated bacteria following the procedure and data given by the system manufacturers. Reagent cards are based on established biochemical methods and developed substrates (23). The results of biochemical reactions were interpreted to establish enzymatic profiles of isolates.

The ability of Bacilli or Actinobacteria to produce antimicrobial agents was tested on standard strains of fungi species of Candida albicans (ATCC^® 10231^TM) and bacterial species of Escherichia coli (ATCC^® 8739^TM), Pseudomonas aeruginosa (ATCC^® 9027^TM), Staphylococcus aureus (ATCC^® 6538^TM), Bacillus subtilis (ATCC^® 6633^TM), Staphylococcus epidermidis (ATCC^® 12228^TM), Klebsiella pneumoniae (ATCC^® 4352^TM), Enterococcus hirae (ATCC^® 10541^TM), vancomycin-resistant Enterococcus faecalis (VRE) (ATCC^® 51299^TM) and methicillin-resistant Staphylococcus aureus (MRSA) (ATCC^® 33591^TM).

Bacterial suspensions containing 3 × 10⁸ cells ml⁻¹ of the selected isolates were prepared. 2.5 μl of each suspension were incubated on Mueller Hinton Agar (MHA) plates at 20°C for 24 h. After incubation, all media in which bacterial colonies were observed, were exposed to ultraviolet (UV) radiation in an open laminar flow cabinet. Therefore, the vitality of the bacteria was destroyed. 1.5 × 10⁸ cells ml⁻¹ of 24 h fresh cultures of the standard strains were prepared. 100 μl of each suspension was mixed with TSA medium at 45°C. Subsequently, it was poured into the previously UV exposed plates, then incubated for 24 hours at 37°C after solidification. At the end of the incubation period, the growth of the standard bacteria in the TSA was investigated and the zone diameters were measured (24).

The sensitivity of 101 selected isolates to antibiotics was examined by using the disc diffusion method of Kirby-Bauer (21) in which 10 antibiotics were used: piperacillin (100 μg), erythromycin (15 μg), vancomycin (30 μg), ampicillin (10 μg), neomycin (10 μg), gentamycin (10 μg), chloramphenicol (30 μg), tetracycline (10 μg), rifampicin (30 μg) and ofloxacin (10 μg). The incubation conditions were 24 h at 20°C. Escherichia coli (ATCC^® 8739^TM), Pseudomonas aeruginosa (ATCC^® 9027^TM) and Staphylococcus aureus (ATCC^® 6538^TM) were tested against the same antibiotics as control microorganisms (25).

Temperature, pH, conductivity and hardness values of water samples are shown in Table I. The air temperature of the sample areas A, B, C (Figure 1) was determined. The temperature of area A was 9.8°C and that of B and C were determined as 9.4°C. The moisture value was evaluated as 93% in all these areas. The Parsık cave resembles most cave systems with its high level of humidity and stable air temperature (26, 27). It was determined that the pH and hardness values of the waters at points B and C were higher than those at point A. These details highlight the differences in chemical environment that may exist within the cave areas.

The highest vitality percentage of bacteria isolated in soil samples was found in samples from point B with 38.7%, whereas the highest vitality percentage in the water samples was found in samples from point C with 26.3% (Table II). In cave environments, it is observed that bacteria can survive metabolically but cannot be cultured. This is because bacteria enter a viable but nonculturable cell form under extreme environmental conditions such as low or high temperature, nutrient deficiency, osmolarity and light. In addition, cave microorganisms obtain their energy from the cave atmosphere or the cave surfaces to which they are attached (28, 29).

SCA, ISP4, SEA and AIA-G have been used especially in surface and soil samples to increase the probability of isolating bacteria belonging to phylum Actinobacteria, which have an extremely high potentials in terms of antimicrobial production (30). TWA and R2A medium were used for both isolation and counting of other bacterial groups. Apart from these media, 1/2 TSA was used for isolation of other bacterial groups from all samples. The cave environment in general is oligotrophic and these media provide a similar environment to the culturable cave bacteria. The number of culturable aerobic heterotrophic bacteria from water and soil samples obtained from R2A and TWA media is given in Figure 2.

Fig. 2

When the bacterial counts of water and soil samples in R2A and TWA media were examined, the highest bacterial numbers were found in R2A medium. These results were evaluated statistically using the Kruskal-Wallis test. The p value was found to be 0.09 and no significant difference was found between the numbers of bacteria grown on the R2A and TWA media. In a study conducted in 2014 (31), the efficiency of various media (SEA, TWA, SCA, TSA) was compared to their suitability for bacterial counting. Efficient results for both isolation and counting were obtained in TWA.

A total of 372 bacteria were isolated from all samples. VITEK^® analysis was applied to only 321 bacteria which had different characteristics in culture-based analyses. The results of the systematic classification of the bacteria were compiled by biochemical analysis using the VITEK^® 2 Compact 30 automated system. Actinobacteria (33%) was determined to be the dominant phylum in this study while the other determined phyla were Firmicutes (25%) and Proteobacteria (16%). In our previous work in Kadıini cave in Turkey, the dominant phylum was Firmicutes (86%), followed by Proteobacteria (12%) and Actinobacteria (2%) respectively (32). In addition, in the study done by Tomova et al. (33), Proteobacteria (51.45%) were found to be the dominant phylum in the samples taken from the Magura cave, Bulgaria, followed by Actinobacteria (43.68%) and Bacteroidetes (3.88%). Although the bacterial habitat of each cave is specific, Proteobacteria, Actinobacteria, Bacteroidetes and Firmicutes are the most identified groups in culture-based microbiological studies in caves (34–36).

In our study, Firmicutes was the most common phylum in soil samples with a rate of 33%, while the most common phylum determined in surface and water isolates was Actinobacteria with 36% and 35% respectively. Considering all the samples, at the class level, Actinobacteria was the most dominant with 33%, while Bacilli with 23% was detected as the second dominant class. It was demonstrated through previous studies that Actinobacteria existed mainly in cave walls, soil, sediment and on speleothem surfaces, which might have considerably contributed to the formation of cave structures and the biomineralisation in the cave ecosystems (4–37). Actinobacteria as well as Firmicutes are frequent among the microbial population inhabiting the caves. Compared to the Proteobacteria group, Firmicutes are more resistant to stress caused by dehydration as well as restriction of nutrients (37). Contrary to our findings for Parsık cave, Proteobacteria are a dominant group in heterotrophic bacterial communities in many caves (33, 34, 38–40). In the current study, Proteobacteria were determined respectively as 10%, 21% and 17% in the surface, water and soil samples. The dominant classes of this phylum were found to be Gammaproteobacteria and Alphaproteobacteria with 9.2% and 6.4%, respectively. In our previous study in Kadıini cave, Alphaproteobacteria were detected at 2%, while Gammaproteobacteria were at 9% (32). The phylum Proteobacteria, having a key role in biogeochemical cycles, and being abundant in samples from cave sediment, soil, dripping water and cave surface, is a cosmopolitan bacterial group (37).

Enzymatic reactions of microorganisms give us ideas of their metabolic activities which are related to their environment. The biochemical tests of our isolates in the VITEK^® system were not only useful for bacterial identification but also to provide more information about nutrients in Parsık cave. In addition, results of these tests were used to evaluate the potentials of the isolates for biotechnological uses in terms of their enzyme production. 76 Gram-negative bacteria, 194 Gram-positive and 51 Gram-positive spore forming bacteria have been tested using the GN, GP and BCL cards respectively in the VITEK^® 2 compact device, and results are given in Figure 3, Figure 4 and Figure 5 respectively. Most of the isolates displayed peptidase (arylamidase) while only Gram-negative bacteria (less than 10%) showed lipolytic activity. In the study conducted in Gumki cave, India, 75.5% of bacteria produced lipase, 47% were amylase producers and 24% produced protease (41). Another study screening cave bacteria for enzyme production found 40% lipase and 87% protease producers (33). This variation in enzymatic profiles in cave bacteria reinforces the idea that every cave is unique.

Fig. 3

Fig. 4

Fig. 5

The high activity of amino acids arylamidase determined in our tested isolates indicates their potential for protein catabolism (42). The phyla Firmicutes (31%) and Actinobacteria (30.7%) produced the highest amounts of arylamidases identified among the tested isolates. 85.52%, 65.97% and 82.35% of Gram-negative, Gram-positive and Gram-positive spore forming bacilli revealed tyrosine-arylamidase activity. Tyrosine is a non-essential amino acid which is synthesised through phenylalanine hydrolysis. It plays a major role in most enzyme synthesis as reported by Kalkan and Altuğ (42), since it is the phosphate and sulfate receptor of protein kinase during protein synthesis. It is also used to reinforce the activity of proteins as demonstrated in a study conducted in thrombin inhibitors showing that tyrosine sulfation could open a way for the development of an anti-thrombotic drug (43). Hence, tyrosine arylamidase has a valuable role in biotechnology since it contributes to the liberation of the amino acid tyrosine.

Enzymes like leucine arylamidase have been reported to be important in food processing industries and the treatment of waste products (44, 45). The degradation of leucine and other amino acids results in volatile molecules responsible for the flavours of some foods like meat products as reported by Papamanoli et al. (46) and Lee et al. (44). In addition, a study showed the roles of bacteria in conversion of paper mill sludge demonstrating the important contribution of amino acid peptidase with leucine arylamidase (45). In our study, 81.95% and 88.23% of Gram-positive bacteria and Gram-positive bacilli showed leucine arylamidase activity. This enzyme was the second most produced enzyme, after the tyrosine arylamidase, by our isolates. Bacteria which can produce this enzyme could be used directly or indirectly by using their enzymes in both composting of sludge and fermentation of food products such as meat and dairy products.

VITEK^® results have showed that some Parsık cave isolates exhibit beta-galactosidase activity which is the more expressed carbohydrate hydrolase in this study. Considering the whole of the tested isolates, most of the bacteria producing beta-galactosidase belong to the Firmicutes phylum with 40.6%, while only 10.9% of beta-galactosidase producers were classified under the phylum Proteobacteria.

The main role of the beta-galactosidase enzyme is to convert lactose into monosaccharides. Glucose and galactose resulting from this reaction not only contribute to the development of the cell but can also be used in dairy product processing (44, 47). This enzyme is important since it solves the problem of human lactose intolerance. The hydrolysation of lactose by this enzyme results in molecules like galactooligosaccharides which have health benefits as prebiotics (47). Moreover, breakdown of some sugars like D-mannose, D-mannitol and D-glucose by fermentation was reported, especially in Gram-negative bacteria.

Lipolytic activity was also observed in some of our isolated Gram-negative bacteria (less than 10%). Even if it was produced by a minimum number of isolates, the activity of lipase was fully expressed by bacteria belonging to the phylum Proteobacteria. This class of enzymes which is used in hydrolysation of lipids could be important in bioremediation since it could participate in oil degradation. Sharma et al. reported that microbial lipases are best for biodiesel production (48). Since they can use all types of free fatty acids and glycerides, they exhibit a high activity, thermostability, alcohol resistance, less reaction time as well as less production inhibition (48). Other enzymes were produced by some of the bacteria in Parsık caves. Further studies should be carried out to clarify them and assess their biotechnological uses.

Microorganisms with broad-spectrum bioactive components, antifungal and antibacterial agents in cave-specific habitats are common in these extreme environments (17). In our study, a total of 129 cave bacteria were tested for their antimicrobial effect against nine different standard bacterial strains and one fungal strain. Experiments have shown that 10 of the selected bacteria (six from Actinobacteria class, four from Bacilli class) have antimicrobial effects against the standard strains.

Parsık cave isolates displayed variable inhibition rates against the tested microorganisms and the highest inhibition rate was observed against Candida albicans. Some of our cave isolates have been found to have inhibitory effects against S. aureus, S. epidermidis, VRE and P. aeruginosa. The zone diameters of cave bacteria with antimicrobial properties against tested microorganisms are shown in Table III.

In our study, the isolate which affected S. epidermidis belongs to the Bacilli class and those which inhibit VRE and S. aureus belong to the Actinobacteria class. Some studies have shown that bacteria with antimicrobial activity inhabiting karst caves are often from the Actinobacteria class (30, 31). However, cave bacteria belonging to phyla Proteobacteria, Firmicutes (especially Bacilli class) and Bacteroides were determined to have antimicrobial and bioactive substances. Thus, approximately 50% of the bacteria isolated from the Magura cave, Bulgaria were detected to inhibit the increase of P. aeruginosa (33). Cave bacteria inhibiting MRSA and VRE clinical strains were determined in a study on Actinomycetes isolated from 19 different caves in Turkey (30). Certainly the bacteria belonging to the class Actinobacteria are the best known in terms of antimicrobial material synthesis, but the isolation of bacteria belonging to the other classes is very important especially in karst environments.

Antibiotic resistant bacteria are widespread in several environments. In this study, resistance to 10 different antibiotics of 101 bacteria (76% Gram-positive; 25% Gram-negative) selected from the cave isolates was investigated. Isolates with a metabolic reaction rate of at least 95% similarities to the data in the VITEK^® database were selected.

When the antibiotic resistance profiles of the isolates were examined, 7% of the bacteria belonging to the cave were resistant to all antibiotics. The highest number of bacteria showed resistance against ampicillin with a rate of 38.6%. In addition, 35.6% of the isolates showed resistance against two or more antibiotics.

Antibiotic resistance profiles of Gram-positive and Gram-negative cave isolates are shown in Figure 6. The lowest resistance was observed to rifampicin (9% for Gram-positive and 8% for Gram-negative). In parallel with our study, it was determined that all the Pajsarjeva jama, Slovenia, isolates were sensitive to rifampicin (49). Likewise, low levels of resistance to ofloxacin, which is a DNA/RNA synthesis inhibitor like rifampicin, were observed in Parsık cave isolates (11% in Gram-positive and 12% in Gram-negative). The resistance rate of Pajsarjeva jama isolates to erythromycin was 73.6% for Gram-negative and 39% for Gram-positive bacteria. The resistance of Parsik cave isolates to erythromycin was determined at lower levels of 20% and 21% for Gram-negative and Gram-positive bacteria respectively. The levels of resistance to protein synthesis inhibitors other than erythromycin (gentamycin, neomycin, tetracycline and chloramphenicol) were determined to range from 12% to 20% for both Gram-positive and Gram-negative bacteria. Contrary to our study, Lavoie et al. (50) showed that cave isolates were highly resistant to gentamycin, neomycin and chloramphenicol antibiotics (33–66% for Gram-negative bacteria and 61–83% for Gram-positive bacteria).

Fig. 6

Furthermore, the lowest resistance to cell wall synthesis inhibitors was observed in piperacillin for both Gram-positive (12%) and Gram-negative (16%) bacteria. In the study conducted by Lavoie et al. (50), the resistance to piperacillin, compared to other antibiotic resistance, was found to be lower (average 37.5%).

In our study, considering the cell wall synthesis inhibitors (vancomycin and ampicillin), Gram-negative bacteria were found to be more resistant than the Gram-positive ones. Similar to our study, Avguštin et al. (49) revealed that cave Gram-negative isolates showed higher resistance to ampicillin.

According to VITEK^® results, except ampicillin and vancomycin, Actinobacteria were determined to be the most resistant (47–70%) phylum to all tested antibiotics. The highest resistance to ampicillin and vancomycin was observed in the phylum Proteobacteria. Like the microbial diversity of caves, antibiotic resistance is also variable. While the antibiotic resistance rates were high, no isolate producing antimicrobial agent was detected in the study conducted by Lavoie et al. (50). However, one of the antibiotic resistance hypotheses in caves is that there is a high rate of antibiotic resistance in the presence of microorganisms producing antimicrobial agents. Studies have shown that bacteria having antibiotic genes can also produce antimicrobial agents (51, 52). In our study, it was found that 50% of the isolates producing antimicrobial agents were resistant to at least two antibiotics. Therefore, study of bacterial antibiotic resistance may contribute to the development of new antibiotics. To clarify this issue, studies in this issue should be continued.

Although much progress has been made in the use of renewable energy in recent years, fossil fuels (especially oil and gas) still meet most of the global energy demand, and they will continue to be the dominant source of energy worldwide over the next few decades (1).

Petroleum is a naturally occurring material found in various geological formations (reservoirs) worldwide. Crude oil, the liquid part of petroleum, is primarily composed of hydrocarbons (2). However, it may also include compounds of nitrogen, sulfur, oxygen and metals (3). Because crude oil in reservoirs is found as a mixture containing varying constituents and proportions, each crude oil has its own unique properties. The most important specified properties are density and sulfur content (4). The density of crude oil is reported in terms of American Petroleum Institute (API) gravity (specific gravity). Based on the API gravity, crude oils can be classified into light, medium, heavy and extra heavy oils (3). Depending on the amount of sulfur content (elemental sulfur or sulfur compounds such as hydrogen sulfide), the crude oil is categorised as ‘sweet’ or ‘sour’. In addition to chemical composition and physical properties, crude oil typically is also identified by underground reservoir (4). Reservoir characteristics (depth, temperature, pressure and other factors) vary significantly from one location to another, even in the same geologic formation (5, 6). The fact that microbial community composition and reservoir conditions vary dramatically not only between the different geographical areas, but also among different oil fields in the same region, makes each oil reservoir ecosystem unique.

Despite the extreme environmental conditions in the oil-bearing formations (i.e. anoxic, high temperature, high salinity), many microorganisms are capable of surviving in the oil and water phases of the oil wells (7, 8). Oil fields harbour mainly facultative aerobic and strictly anaerobic microorganisms due to the low redox potential in the reservoirs (8). These ecosystems contain different types of microbial communities (such as mesophiles, thermophiles and halophiles) which adapt to the reservoir conditions (9). Bacterial and archaeal groups identified in oil fields include sulfate-reducing bacteria (10), sulfur-oxidising bacteria (11), methanogens (12), fermentative microorganisms (13), acetogens (14), nitrate reducers (15), manganese and iron reducers (16) and hydrocarbon degraders (17). Among these microbes, sulfate-reducing bacteria have attracted much attention due to their detrimental effects such as reservoir souring and biocorrosion (7). In addition, different members of the oil microbial community are involved in syntrophic interactions. Fermenting bacteria and methanogenic archaea are involved in methanogenic hydrocarbon biodegradation through their close syntrophic associations (18). This microbial process is undesirable in oil reservoirs because it causes a decrease in oil quality and value (19). Syntrophic microorganisms in oil reservoirs also play important roles in the global biogeochemical cycling of sulfur, carbon and nitrogen. For instance, sulfate-reducing bacteria and sulfur-oxidising bacteria, the key drivers in sulfur transformations, are involved in the sulfur cycle (11). Thus, knowledge of the microbial groups and microbial dynamics in oil fields enable us to obtain detailed insights into the microbial ecology of oil associated environments.

Understanding the microbial ecology of oil reservoirs is crucial to the petroleum industry because the success of oilfield operations is strongly influenced by the activity of microorganisms. Oil microbes with different metabolic capabilities have significant negative and positive impacts on the petroleum resources and the extraction processes (7). Microbial activity may lead to severe problems such as reservoir souring and microbial corrosion. Reservoir souring, which is characterised by an increase in production of H₂S in the reservoir fluids, most commonly occurs when sulfidogenic microorganisms reduce sulfate to sulfide, a toxic and corrosive product (20). Undesirable accumulation of sulfide minerals in reservoirs is one of the major challenging problems in oil production because it causes plugging of reservoirs, decreasing the oil quality and value and increasing the refining costs. Moreover, exposure to H₂S can be dangerous in terms of worker health and safety due to its high toxicity. Additionally, the produced H₂S promotes corrosion of the metallic equipment and structures used for oil production and processing (21). Another destructive phenomenon is biocorrosion, which is defined as microbial attack on the surface of the metal infrastructure leading to disruption of the material (22). In addition to sulfate-reducing bacteria, which play a major role in biocorrosion, other corrosive microbes, such as acetogenic bacteria and methanogenic archaea, are also associated with corrosion failures (23). Biocorrosion is a great concern because it leads to loss of material, large economic losses and safety issues in the oil industry (24). In contrast, hydrocarbon-degrading bacteria may be used for environmental clean-up processes (6). Bacterial degradation of hydrocarbons was carried out by both aerobic (for example, Rhodococcus sp., Sphingomonas sp., Pseudomonas putida, Pseudomonas stutzeri, Acinetobacter sp.) and anaerobic bacteria (such as Fe(III)-reducing bacteria, sulfate-reducing bacteria) (6, 17). Furthermore, microbial products such as biopolymers and biosurfactants can be used for facilitating oil movement in a widely used technology, known as microbial enhanced oil recovery (MEOR) (1). Compared with other conventional oil recovery techniques, MEOR has advantages such as low cost, wide application, high efficiency and low environmental pollution (25). Therefore, diversity, metabolic processes and habitat conditions of microbial communities in oil reservoirs should be investigated, so that their negative effects can be decreased and their positive effects can be exploited.

This study aimed to determine the bacterial community composition and to identify the predominant community members in produced water from oil fields located in the Diyarbakır region in Turkey. To this end, we used PCR-DGGE to analyse 20 produced water samples from the Diyarbakır region. There are limited studies on produced water from the Diyarbakır region and this paper represents the only in situ study available. The results of this study provide not only new data about the microbial ecology of the Diyarbakır oil fields, but also information on the bacterial populations which may have potential roles in terms of increasing or decreasing the efficiency of industrial applications.

2.1 Sampling Procedure

2.2 DNA Extraction

2.3 Polymerase Chain Reaction Amplification

2.4 Denaturing Gradient Gel Electrophoresis

2.5 Comparative Sequence Analysis

3.1 Molecular Analysis of Bacterial Communities

In order to increase our knowledge about microbial diversity, culture-dependent and molecular-based approaches are used for describing the diversity of microbes. Molecular-based approaches such as PCR-DGGE methodology, which is a useful tool for monitoring the genetic diversity of complex microbial populations (26), provide valuable information about the microbial community structure and dynamics in nature. For these reasons, PCR-DGGE fingerprinting analysis of environmental samples was used in this study.

The choice of appropriate primers for PCR amplification is a crucial step to accurately characterise the microbial communities. In this study, primer pair (341F-GC/907R), targeting the V3-V5 region of the 16S rRNA gene fragment, was selected due to its suitability for DGGE analysis of bacterial populations in environmental samples (26). This primer pair designed by Muyzer et al. (27, 28) has been used predominantly for microbial community analysis (26).

The DNA yield obtained from produced water samples was very low. It is known that crude oil samples contain low amounts of biomass which makes DNA isolation difficult to achieve (46). In this study, the permit included taking up to 500 ml of oil/water mixture from each sampling point so that only ca. 25 ml of each produced water sample could be obtained. In this scope, the low sample volumes of produced water separated from the oil/water mixture may be a reason for the low amount of DNA. It was reported in other studies that higher sample volumes (100–4000 ml) of produced water were used for DNA isolation (35, 47–50). The low DNA yield affected the efficiency of the PCR technique and for this reason, a two-step PCR was applied to the produced water samples. Thus, a sufficient amount of PCR product for DGGE for the produced water samples could be obtained.

Bacterial communities associated with the produced waters was analysed by the PCR-DGGE approach. Although numerous bands were visible on the DGGE gel, only dominant bands could be excised and sequenced. Most of the sequences retrieved from produced water samples were related to unclassified bacteria. Different studies on oil reservoir microbiota have also shown that oil fields harbour new and still unidentified microbial species. For example, Lenchi et al. described microbial communities in production and injection waters from the Algerian oil fields. In their study, they detected that a large number of unclassified bacterial and archaeal sequences were found in the water samples (51). Furthermore, uncultured bacteria such as uncultured Sphingomonas sp. and uncultured Caenispirillum sp. clone Ppss_Ma27 were detected in our study. This result is consistent with the fact that the vast majority of microorganisms are uncultured and do not grow under laboratory conditions as stated by Lewis et al. (52). In order to isolate more microbes, an appropriate identification laboratory protocol should be followed. At this point, different strategies such as mimicking natural conditions via decreased nutrient, extended incubation times, the modification of isolating media formulations and different incubation parameters (for example, temperature) were suggested for the cultivation of microorganisms (53). For instance, pollutant degrader Sphingomonas, which seemed to be previously uncultured by nutrient-rich methods, could be isolated from crude oil contaminated soil by using an in situ method that mimics the original environment (54). In addition, culture-dependent investigation should also be supported by molecular techniques.

Based on the sequences, organisms related to known mesophilic bacteria were predominant in the produced water samples. In addition, some organisms related to thermophilic bacteria (Aeribacillus pallidus, Marinobacter sp. Trimyema-2, Phenylobacterium lituiforme and Thermithiobacillus sp.) were also identified. Bacteria having different metabolic capabilities (denitrifying, biodegrading and sulfur removing bacteria) were also detected. In addition, bacteria which may cause biocorrosion on steel surfaces were detected.

The dominant bacterial phylum was the Proteobacteria. The members of this phylum were also frequently found in many other studies on microbial diversity of oil field produced waters (55–58). Moreover, it was stated that Proteobacteria are ubiquitous in oil reservoirs over all temperature ranges (59).

In this study, among the detected genera in produced water samples that potentially contain hydrocarbon degrading bacteria were Aeribacillus, Acinetobacter, Sphingomonas, Marinobacter and Phenylobacterium. It has been known for years that the species belonging to these genera are capable of degrading hydrocarbons (6, 17, 60, 61). In addition, G. daeguensis, a hydrocarbonoclastic bacterium, was detected in produced water sample with a 100% sequence similarity. G. daeguensis has also been demonstrated as a potential microbe for bioremediation due to its hydrocarbon degradation ability (44). Further investigations are needed because our current knowledge of the metabolic capability of G. daeguensis is limited. Moreover, sulfur-oxidising Thermithiobacillus sp. was also identified in produced water sample. Sulfur-oxidising bacteria, which oxidise the sulfur compounds produced by the activity of sulfate-reducing bacteria in oil reservoirs, may play a key role in the oil industry because they can be utilised to resolve processing problems such as reservoir souring (11).

Pseudomonas was the dominant genus detected among the produced water samples. Pseudomonas stutzeri was the species identified in five produced water samples. P. stutzeri was previously isolated not only from formation water, produced from the petroleum wells in Adıyaman (62), but also oil-contaminated soils in Batman petroleum refinery, Turkey (63). These two areas are close to the Diyarbakır region from where the samples in this study were collected and these findings show that P. stutzeri is distributed widely in south-eastern Turkey. In other different geographical areas, this species was also isolated from oil-associated environments, such as oil field production water (64), oil sludge (65) and oil contaminated soil (66). However, although P. stutzeri is often isolated from oil reservoirs, the origin of P. stutzeri in oil reservoirs is a debatable issue. Because oil reservoirs have low redox potentials and contain little oxygen, anaerobic microorganisms are considered as truly indigenous to oil reservoirs (67). In this regard, it is believed that P. stutzeri, most of whose strains are aerobes, is an exogenous organism inoculated into oil reservoirs during the oil production processes. Even if strains of P. stutzeri are introduced into oil reservoirs with injected fluids, they should adapt to the physicochemical characteristics of the reservoir to survive. At this point, it has been proposed that extreme reservoir conditions may act as special factors for the evolution of P. stutzeri, thereby forming mutant strains (68). Furthermore, P. stutzeri, being found in a wide variety of habitats, is known for its diverse metabolism. Some strains of P. stutzeri are capable of denitrification, degradation of aromatic compounds and nitrogen fixation (32). These metabolic features make P. stutzeri highly attractive for biotechnological processes, such as reservoir souring control (69), microbial enhanced oil recovery (64) and bioremediation of oil-polluted environments (65).

In undisturbed oil reservoirs, microorganisms are found in different phases such as reservoir fluid containing crude oil and formation water, and rock surfaces. While planktonic microbes thrive in the water phase, sessile microbes may attach to oil or rock surfaces (59). In addition, biofilm may form on the metal surfaces of the pipes in the oil-producing wells (70). Oil microbiome studies focus mainly on the analysis of the water phase due to its easy sampling. However, it should be noted that the water phase itself contains only a minor portion of the microbes found in the oil reservoir (59). On the other hand, the sampling of sessile microbes is likely to be more challenging (70).

This study reported for the first time the bacterial community composition of produced water from Diyarbakır oil reservoirs as obtained by DGGE analysis of PCR-amplified 16S rRNA gene fragments. DGGE analysis of produced water samples demonstrated that the majority of the bacterial sequences belonged to unclassified bacteria, indicating that oil reservoirs harbour still undescribed microbial species. Among the classified bacteria, the members of Proteobacteria were more abundant. Pseudomonas was the dominant genus detected in the produced water. Although the members of Pseudomonas were known as exogenous organisms inoculated into oil reservoirs, Pseudomonas stutzeri was found in five produced water samples. Bacteria having different metabolic capabilities (denitrifying, biodegrading and sulfur removing bacteria) were also detected. It can be stated that the metabolic capacities of these bacteria make them potential candidates for utilising in biodegradation, bioremediation, the improvement of oil quality and oil recovery processes. The knowledge of the bacterial community composition in oil reservoirs of the Diyarbakır region obtained in this study will be of great interest for both scientific research and applications in the oil industry. To build on the data presented in this study, metagenomic analyses should be performed to explore the undescribed microbes.