Search

Monitoring environmental status through molecular investigation of microorganisms in the marine environment is suggested as a potentially very effective method for biomonitoring, with great potential for automation. There are several hurdles to that approach with regards to primer design, variability across geographical locations, seasons, and type of environmental pollution. Here, qPCR analysis of genes involved in the initial activation of aliphatic and aromatic hydrocarbons were used in a laboratory setup mimicking realistic oil leakage at sea. Seawater incubation experiments were carried out under two different seasons with two different oil types. Degenerate primers targeting initial oxygenases (alkane 1-monooxygenase; alkB and aromatic-ring hydroxylating dioxygenase; ARHD) were employed in qPCR assays to quantify the abundance of genes essential for oil degradation. Shotgun metagenomics was used to map the overall community dynamics and the diversity of alkB and ARHD genes represented in the microbial community. The amplicons generated through the qPCR assays were sequenced to reveal the diversity of oil-degradation related genes captured by the degenerate primers. We identified a major mismatch between the taxonomic diversity of alkB and ARHD genes amplified by the degenerate primers and those identified through shotgun metagenomics. More specifically, the designed primers did not amplify the alkB genes of the two most abundant alkane degraders that bloomed in the experiments, Oceanobacter and Oleispira. The relative abundance of alkB sequences from shotgun metagenomics and 16S rRNA-based Oleispira-specific qPCR assay were better signals for oil in water than the tested qPCR alkB assay. The ARHD assay showed a good agreement with PAHs degradation despite covering only 25% of the top 100 ARHD genes and missing several abundant Cycloclasticus sequences that were present in the metagenome. We conclude that further improvement of the degenerate primer approach is needed to rely on the use of oxygenase-related qPCR assays for oil leakage detection. publishedVersion

Environmental DNA (eDNA) is a novel, non-invasive sampling procedure that allows the obtaining of genetic material directly from environmental samples without any evidence of biological sources. The eDNA methodology can greatly benefit from coupling it to reliable, portable and cost-effective tools able to perform decentralized measurements directly at the site of need and in resource-limited settings. Herein, we report a simple method for the selective analysis of eDNA using a magneto-assay with electrochemical detection. The proposed method involves the polymerase chain-reaction (PCR) amplification of mitochondrial eDNA of parasitic Salmon lice (Lepeophtheirus salmonis), extracted from seawater samples. The eDNA sequence was targeted via sandwich hybridization onto magnetic beads and enzymatic labeling was performed to obtain an electroactive product measured by differential pulse voltammetry. Quality by Design (QbD), a recent concept of science- and risk-oriented quality paradigm, was used for the optimization of the different parameters of the assay. Response surface methodology and Monte Carlo simulations were performed to define the method operable design region. The optimized electrochemical magneto-assay attained a limit of detection of 2.9 amol μL−1 of the short synthetic sea louse DNA analogue (43 bp). In addition, robustness testing using a further experimental design approach was performed for monitoring eDNA amplicons. Seawater samples spiked with individuals of free-swimming L. salmonis copepodite stages and seawater collected from tanks with sea lice-infested fish were analyzed.

The use of natural marine bacteria as &ldquo, oil sensors&rdquo, for the detection of pollution events can be suggested as a novel way of monitoring oil occurrence at sea. Nucleic acid-based devices generically called genosensors are emerging as potentially promising tools for in situ detection of specific microbial marker genes suited for that purpose. Functional marker genes are particularly interesting as targets for oil-related genosensing but their identification remains a challenge. Here, seawater samples, collected in tanks with oil addition mimicking a realistic oil spill scenario, were filtered and archived by the Environmental Sample Processor (ESP), a fully robotized genosensor, and the samples were then used for post-retrieval metatranscriptomic analysis. After extraction, RNA from ESP-archived samples at start, Day 4 and Day 7 of the experiment was used for sequencing. Metatranscriptomics revealed that several KEGG pathways were significantly enriched in samples exposed to oil. However, these pathways were highly expressed also in the non-oil-exposed water samples, most likely as a result of the release of natural organic matter from decaying phytoplankton. Temporary peaks of aliphatic alcohol and aldehyde dehydrogenases and monoaromatic ring-degrading enzymes (e.g., ben, box, and dmp clusters) were observed on Day 4 in both control and oil-exposed and non-exposed tanks. Few alkane 1-monooxygenase genes were upregulated on oil, mostly transcribed by families Porticoccaceae and Rhodobacteraceae, together with aromatic ring-hydroxylating dioxygenases, mostly transcribed by Rhodobacteraceae. Few transcripts from obligate hydrocarbonoclastic genera of Alcanivorax, Oleispira and Cycloclasticus were significantly enriched in the oil-treated exposed tank in comparison to control the non-exposed tank, and these were mostly transporters and genes involved in nitrogen and phosphorous acquisition. This study highlights the importance of seasonality, i.e., phytoplankton occurrence and senescence leading to organic compound release which can be used preferentially by bacteria over oil compounds, delaying the latter process. As a result, such seasonal effect can reduce the sensitivity of genosensing tools employing bacterial functional genes to sense oil. A better understanding of the use of natural organic matter by bacteria involved in oil-biodegradation is needed to develop an array of functional markers enabling the rapid and specific in situ detection of anthropogenic pollution.

Today's benthic offshore biological monitoring of oil & gas (O&G) activities relies on macrofauna taxa enumeration. For the future, analysis of DNA isolated directly from sediments holds great potential for multi-trophic biodiversity surveys and the monitoring of a larger spectrum of benthic taxa, including micro-fauna. Here, we evaluate more specifically the potential of microfauna-specific gene quantification in relation to both petroleum-related discharge compounds and other seafloor environmental properties. We carried out this evaluation using sediment samples collected at drilling Region III on the Norwegian continental shelf where DNA metabarcoding of eukaryotic diversity was already performed. Generally, the quantification of microfauna indicator taxa related well to the gradient of contamination on the seafloor. Contrary to eukaryotic Euplotida, metabarcoding data and qPCR numbers for indicative prokaryotic taxa showed the same relationship to offshore contaminants (both showed positive relationship). We found absolute numbers of SSU rRNA gene copies of (1) Dinophyceae, Bacillariophyceae and Alcanivorax were correlated with the level of petroleum-related compounds but not with other environmental variables, (2) bacteria closely related to Shewanella were correlated with the concentration of Ba, PAH, as well to percent of gravel, (3) Desulfobacteriales correlated with petroleum-related contaminants, but as well with percent of gravel and grain size. Findings from our study suggest that biomonitoring surveys of O&G activities on benthos could benefit from quantification of specific micro-fauna indicators that is simpler and faster than the methods currently used for impact assessment of benthos.

In this paper, a molecular analytical approach for detecting hydrocarbonoclastic bacteria in water is suggested as a proxy measurement for tracking petroleum discharges in industrialized or pristine aquatic environments. This approach is tested for general application in cold marine regions (freezing to 5 °C). We used amplicon sequencing and qPCR to quantify 16S rRNA and GyrB genes from oleophilic bacteria in seawater samples from two different crude oil enrichments. The first experiment was conducted in a controlled environment using laboratory conditions and natural North Sea fjord seawater (NSC) at a constant temperature of 5 °C. The second was performed in the field with natural Arctic seawater (ARC) and outdoor temperature conditions from −7 °C to around 4 °C. Although the experimental conditions for NSC and ARC differed, the temporal changes in bacterial communities were comparable and reflected oil biotransformation processes. The common bacterial OTUs for NSC and ARC had the highest identity to Colwellia rossensis and Oleispira antarctica rRNA sequences and were enriched within a few days in both conditions. Other typical oil degrading bacteria such as Alcanivorax (n-alkane degrader) and Cycloclasticus (polycyclic aromatic hydrocarbons degrader) were rapidly enriched only in NSC conditions. Both the strong correlation between Oleispira SSU gene copies and oil concentration, and the specificity of the Oleispira assay suggest that this organism is a robust bioindicator for seawater contaminated by petroleum in cold water environments. Further optimization for automation of the Oleispira assay for in situ analysis with a genosensing device is underway. The assay for Colwellia quantification requires more specificity to fewer Colwellia OTUs and a well-established dose-response relationship before those taxa are used for oil tracking purposes.

For the environmental monitoring of coral, mucus appears to be an appropriate biological matrix due to its array of functions in coral biology and the non-intrusive manner in which it can be collected. The aim of the present study was to evaluate the feasibility of using mucus of the stony coral Lophelia pertusa (L. pertusa) as an analytical matrix for discovery of biomarkers used for environmental monitoring. More specifically, to assess whether a mass-spectrometry-based proteomic approach can be applied to characterize the protein composition of coral mucus and changes related to petroleum discharges at the seafloor. Surface-enhanced laser desorption/ionization-time of flight mass spectrometry (SELDI-TOF MS) screening analyses of orange and white L. pertusa showed that the mucosal protein composition varies significantly with color phenotype, a pattern not reported prior to this study. Hence, to reduce variability from phenotype difference, L. pertusa white individuals only were selected to characterize in more detail the basal protein composition in mucus using liquid chromatography, mass spectrometry, mass spectrometry (LC-MS/MS). In total, 297 proteins were identified in L. pertusa mucus of unexposed coral individuals. Individuals exposed to drill cuttings in the range 2 to 12 mg/L showed modifications in coral mucus protein composition compared to unexposed corals. Although the results were somewhat inconsistent between individuals and require further validation in both the lab and the field, this study demonstrated preliminary encouraging results for discovery of protein markers in coral mucus that might provide more comprehensive insight into potential consequences attributed to anthropogenic stressors and may be used in future monitoring of coral health.

Ocean acidification (OA) resulting from anthropogenic emissions of carbon dioxide (CO(2)) has already lowered and is predicted to further lower surface ocean pH. There is a particular need to study effects of OA on organisms living in cold-water environments due to the higher solubility of CO(2) at lower temperatures. Mussel larvae (Mytilus edulis) and shrimp larvae (Pandalus borealis) were kept under an ocean acidification scenario predicted for the year 2100 (pH 7.6) and compared against identical batches of organisms held under the current oceanic pH of 8.1, which acted as a control. The temperature was held at a constant 10°C in the mussel experiment and at 5°C in the shrimp experiment. There was no marked effect on fertilization success, development time, or abnormality to the D-shell stage, or on feeding of mussel larvae in the low-pH (pH 7.6) treatment. Mytilus edulis larvae were still able to develop a shell in seawater undersaturated with respect to aragonite (a mineral form of CaCO(3)), but the size of low-pH larvae was significantly smaller than in the control. After 2 mo of exposure the mussels were 28% smaller in the pH 7.6 treatment than in the control. The experiment with Pandalus borealis larvae ran from 1 through 35 days post hatch. Survival of shrimp larvae was not reduced after 5 wk of exposure to pH 7.6, but a significant delay in zoeal progression (development time) was observed.

Polar marine surface waters are characterized by high levels of dissolved oxygen, seasonally intense UV irradiance and high levels of dissolved organic carbon. Therefore, the Arctic sea-ice habitat is regarded as a strongly pro-oxidant environment, even though its significant ice cover protects the ice-associated ( = sympagic ) fauna from direct irradiation to a large extent. In order to investigate the level of resistance to oxyradical stress, we sampled the sympagic amphipod species Gammarus wilkitzkii during both winter and summer conditions, as well as exposed specimens to simulated levels of near-natural and elevated levels of UV irradiation. Results showed that this amphipod species possessed a much stronger antioxidant capacity during summer than during winter. Also, the experimental UV exposure showed a depletion in antioxidant defences, indicating a negative effect of UV exposure on the total oxyradical scavenging capacity. Another sympagic organism, Onisimus nanseni , was sampled during summer conditions. When compared to G. wilkitzkii , the species showed even higher antioxidant scavenging capacity.

The sensitivity of different tissues for assessment of chronic low-dose environmental exposure of fish to alkylphenols (APs) was investigated. We exposed Atlantic cod (Gadus morhua) in the laboratory to tritium labelled 4-tert-butylphenol, 4n-pentylphenol, 4n-hexylphenol, and 4n-heptylphenol via seawater (8 ng/l) and via contaminated feed (5 microg/kg fish per day). Measurements of different fish tissues during eight days of exposure and eight subsequent days of recovery revealed that APs administered via spiked seawater were readily taken up whereas the uptake was far less efficient when APs were administered in spiked feed. AP residues were mainly located in the bile fluid whereas the concentrations in liver were very low, indicating a rapid excretion and the liver-bile axis to be the major route of elimination. The biological half-life of APs in the exposed cod was short, between 10 and 20 h. Our study shows that in connection with biomonitoring of AP exposure in fish, assessment of AP metabolites in bile fluid is a more sensitive tool than detection of parent AP levels in liver or other internal tissues.

The scientific goal of the present work is the development of a new molecular biomarker as a biomonitoring tool for early assessment of the biological effect of complex mixtures of pollutants (especially polycyclic aromatic hydrocarbons) and validation of the approach by comparison with standard methodologies. Hence, a DNA‐based biosensor, which is able to detect the presence of compounds with affinity for DNA to measure their effect on the guanine base, was used in this research.

Despite the extensive transport of chemicals at sea, there is current lack of knowledge of the fate and effects of many of them on the marine biota. The current regulation that follows the GESAMP-MARPOL classification is mainly based on ecotoxicity assessment from fresh water based studies. Repetitive spills in marine coastal environment from tanker ship loaded with several thousand tonnes of chemicals raised concern about whether the existing freshwater data location can be used to predict the behaviour and the environmental effects of contaminants in marine surroundings. There is a general lack of information of the fate of chemicals at sea. A deviating pattern in marine environment from that in freshwater may have significant consequences for the counteracting actions taken to fight the spill, on staff working on the site of spill as well as on marine life present in the vicinity of the accident. In the present article, an environmental effect study of styrene was conducted as part of the ECOPEL program. We report some biological effects of styrene in laboratory-exposed marine organisms. Styrene was continuously supplied at a nominal concentration of 2mg L(-1) over 7 days to both mussels (Mytilus edulis) and fish (Symphodus mellops). At the end of this period, DNA damage was assessed by the Comet assay performed on blood (fish) and haemolymph (mussel) cells. In mussels, the lysosomal membrane stability was additionally assessed by the neutral red retention time assay (NRRT). Significant biological responses were observed over the studied period in both organisms with these two tests. Hence, the results favour the use of a biomarker-based approach to assess the health conditions in case of spill.

Biomarkers have the potential to be used to assess the impact of anthropogenic discharges in marine waters. We have used a suite of biomarkers spanning from enzymatic to histopathological alterations and general stress responses to assess the short- and long-term impact on mussels Mytilus edulis of heavy fuel oil no. 6 and styrene. Mussels were exposed for 5 months, with a refilling of the exposure system, to a water soluble fraction of heavy fuel and, then, kept for a month in clean water for recovery. In a second experiment, mussels were exposed to styrene for 19 days and maintained in clean water for up to 4 months. Chemical body tissue levels reflected the weathering processes of these compounds. Acyl-CoA oxidase activity was induced in oil-exposed mussels after refilling, whereas styrene inhibited it after 19 days of exposure and after 2 weeks in clean water. Gamete development and alkali-labile phosphate levels suggest that neither oil nor styrene behaved as endocrine disruptors. Neutral red retention time was lower in treated groups than in controls. Lysosomal membrane stability was significantly reduced in exposed groups and recovered after withdrawal of oil but not after removal of styrene. Neither oil nor styrene exposure affected the condition index except for the reduction seen in mussels exposed to oil for 1 month. Biomarker response index discriminated exposed mussels, which showed higher values, and returned to control levels after recovery. Results obtained from these pilot experiments can help to identify relevant monitoring tools to assess the impact of oil and chemicals in marine spill scenarios.

Within the frame of a large environmental study, we report on a research program that investigated the potential for bioaccumulation and subsequent effect responses in several marine organisms exposed to chronic levels of dispersed crude oil. Body burden can be estimated from kinetic parameters (rate constants for uptake and elimination), and appropriate body burden-effect relationships may improve assessments of environmental risks or the potential for such outcomes following chronic discharges at sea. We conducted a series of experiments in a flow-through system to describe the bioaccumulation kinetics of polycyclic aromatic hydrocarbons (PAH) at low concentrations of dispersed crude oils. Mussels (Mytilus edulis) and juvenile turbot (Scophthalmus maximus) were exposed for periods ranging from 8 to 21 d. Postexposure, the organisms were kept for a period of 9 to 10 d in running seawater to study elimination processes. Rate constants of uptake (k1) and elimination (k2) of the PAHs during and following exposure were calculated using a first-order kinetic model that assumed a decrease of the substances in the environment over time. The estimated bioconcentration factor was calculated from the ratio of k1/k2. The kinetic parameters of two-, three-, and four-ring PAHs in mussel and fish are compared with estimates based on hydrophobicity alone, expressed by the octanol-water partition coefficient, Kow (partitioning theory). A combination of reduced bioavailability of PAHs from oil droplets and degradation processes of PAHs in body tissues seems to explain discrepancies between kinetic rates based on Kow and actual kinetic rates measured in fish. Mussels showed a pattern more in compliance with the partitioning theory.

The fate of hydrocarbons discharged at sea was studied in laboratory exposures examining the uptake of PAC (polyaromatic compounds) in marine organisms. Methodology and species selection were guided by the Northwest European (OSPAR) convention. The acute toxicity of poorly soluble substances is normally assessed by preparing water accommodated fractions (WAF). Following such tests we studied PAC uptake in unicellular algae, crustaceans and fish. Preparation of WAF was first improved because in the standard procedure, oil droplet size increased with oil loading and flask size. At similar exposure levels the proportion of large molecular weight PAC associated with algae increased when the density of algae decreased. However, as body size increased via crustaceans to fish more of the smaller molecular weight PAC were bioconqentrated. Organism density and type of organism thus influence the abundance of PAC in food chains.

Biomarkers are widely used to measure environmental impacts on marine species. For many biomarkers, it is not clear how the signal levels relate to effects on the whole organism. This paper shows how species sensitivity distributions (SSDs) can be applied to evaluate multiple biomarker responses in species assemblages. To our knowledge, the present study compared for the first time SSDs based on biomarker response levels for marine species to a SSD for whole-organism responses. The comparison indicates that for exposure to dispersed oil in the marine environment, the selected biomarkers were, on average, 35- to 50-fold more sensitive than the whole-organism effect. At the 5% hazardous concentration derived from the SSD for whole-organism effects, which is a conservative threshold level, the potentially affected fraction of species showing biomarker response corresponds to approximately 80%. Variation in species sensitivity, expressed either as biomarker or as whole-organism response levels, were similar. Although uncertainties exist, the link between biomarkers and risk assessment presented here provides a preliminary guideline for deciding when biomarker responses likely are hazardous and, therefore, require further investigation. © 2009 SETAC.

Sheepshead minnows (Cyprinodon variegatus) were continuously exposed to two concentrations of polycyclic aromatic hydrocarbons (PAHs) dissolved in seawater (sigma PAH = 7.57 and 72.31 microg/L) for 36 d, followed by 8 d of depuration. The PAHs studied were naphthalene (NPH or C0-NPH), phenanthrene (PHE or C0-PHE), pyrene (PYR), 2-methylnaphthalene (C1-NPH), 1,3-dimethylnaphthalene (C2-NPH), 2-isopropylnaphthalene (C3-NPH), 9-methylphenanthrene (C1-PHE), and 9-ethylphenanthrene (C2-PHE). Uptake rate constants (k1) for NPHs increased with increasing degree of alkylation and log value of the octanol/water partition coefficient (Kow), whereas k1 values for three- and four-ring PAHs were lower despite their high log Kow values. Elimination rate constants (k2) for the homologue series of NPHs and PHEs generally increased with decreasing degree of alkylation and log Kow values. However, the depuration time did not directly correlate with the molecular size for nonalkylated PAHs. Bioconcentration factors (BCFs) were estimated from the ratio of k1 to k2 and also directly from PAH concentrations in fish tissue and water samples, and the factors generated by the two methods were very similar. A significant positive correlation was determined between log BCFs and log Kow values for the series of C0- through C3-NPH at both low (r2 = 0.985, p = 0.0077) and high (r2 = 0.956, p = 0.022) exposures, although this correlation was not determined for all the PAHs studied. As a result of increased metabolism and/ or reduced bioavailability with increasing lipophilic character, the estimated BCFs for C0- through C2-PHE and PYR were generally lower than those for C0- through C3-NPH. The two exposure levels revealed minor variations in k1 and k2 values for parent PAHs and in the temporal pattern of PAH metabolite concentrations in bile. The present results indicate that the presence and nature of alkyl groups have a significant influence on the kinetics and metabolism of PAHs in fish.

cited By 22; International audience; Polycyclic aromatic hydrocarbons (PAH) are responsible for a number of adverse effects on aquatic biota due to their persistence and potential for accumulation. The recognition and quantification of biliary PAH metabolites from exposed fish represents a valuable tool in the evaluation of warning signals indicative of exposure. In this work, an electrochemical DNA biosensor is proposed as a screening device for the rapid detection of PAH metabolites in fish bile samples. Preliminary studies were performed using the 3-hydroxy-benzo(a)pyrene as a model compound. The effect of the 3-hydroxy-benzo(a)pyrene on the surface-confined DNA was found to be linearly related to the concentration of the PAH derivative in solution. Then, the DNA biosensor was used to investigate the presence of PAH metabolites in fish bile samples. The sensor was able to clearly distinguish PAH-exposed fish bile samples from unexposed control samples. A total analysis time of 7 min was established (including electrode surface pre-treatment, dsDNA immobilization and interaction steps) and only 10 μL of diluted sample solution were needed. DNA biosensors are cost effective and sensitive; moreover, the coupling of a portable and battery-operating potentiostat with disposable carbon strips meets the requirements for in situ measurements.

Assessing the fate in marine biota of hydrocarbons derived from oil particles that are discharged during exploration and production is of relevant environmental concern. However, a rather complex experimental setup is required to carry out such investigations. In this study, a sophisticated tool, the continuous-flow system (CFS), was used to mimic dispersed oil exposure to marine biota. Polycyclic aromatic hydrocarbon (PAH) uptake was studied in two species, the blue mussel Mytilus edulis and juvenile of the turbot Scophthalmus maximus, and in semipermeable membrane devices (SPMD) exposed to crude oil dispersed in a flow-through system. After an exposure period of 8 to 21 d, elimination in organisms and devices was analyzed for 9 to 10 d following transfer to PAH-free seawater. Principal component analysis (PCA) revealed different PAH patterns. In mussel and SPMD, the PAH profiles were very close to that analyzed in seawater. Slight differences were, however, indicated for large molecules with log Kow above six. Nonachievement of steady-state concentration and bioavailability of PAH in oil droplets may account for these differences. The PAH composition in fish revealed only congeners with two to three aromatic rings. A combination of bioavailability and efficient metabolism of the larger PAH molecules may explain this pattern. The CFS made possible a better understanding of some critical factors governing bioconcentration in marine biota from dispersed oil. Yet the results illustrate that uptake of PAH from exposure to oil particles is complex and that different species may bioconcentrate different molecules depending on factors like life style and metabolic capability to degrade the potential harmful substances. Hence, risk assessment of the actual impact of discharges to marine biota should consider these essential biological and ecological factors.

In order to perform environmental risk assessments with regard to oil contamination in the sea, it is important to obtain knowledge about threshold levels for possible adverse effects in marine organisms. With this objective in mind, selected biomarkers were studied in Atlantic cod (Gadus morhua) chronically exposed to mechanically dispersed crude oil. The fish were exposed for 30 days in a continuous flow system to nominal concentrations of 0.06, 0.25 and 1 ppm North Sea crude oil. Fish were sampled five times during the exposure period. In addition, the 1 ppm group and the control group were sampled 1 week after the end of exposure. Polyaromatic hydrocarbon (PAH) concentrations in the seawater were analysed regularly by direct fluorescence and, at one occasion, by gas chromatography with mass spectrographic detection (GC/MS) measurements. Liver samples were analysed for parent PAH levels by means of GC/MS measurements, and PAH metabolites in bile were analysed by means of fixed wavelength fluorescence. Cytochrome P450 induction in liver was estimated by ethoxyresorufin-O-deethylase (EROD) activity, and hepatic DNA adducts were analysed by the 32P-postlabelling assay. The parent PAH concentrations in liver showed peak levels 3 days after the start of exposure, followed by a reduction towards the end of the experiment. In contrast, the PAH metabolites in bile and EROD activity showed generally increasing levels throughout the whole exposure period, indicating an increased biotransformation efficiency. The level of DNA adducts in the 1 ppm group showed a stable increase during the entire exposure period. Only a slight, non-significant decrease in DNA adduct levels was observed after 7 days of recovery in clean water. Exposure-dependent responses were observed for all three biomarkers. The lowest nominal concentration of dispersed oil in water, 0.06 ppm, corresponded to a measured total PAH concentration in the water of 0.3 ppb. Atlantic cod exposed to this concentration showed increased levels of PAH metabolites in bile and a slight induction of CYP1A, as well as formation of DNA adducts when compared with control fish. Particularly noteworthy is the detection of DNA adducts at such a low exposure concentration of oil in water, which, to our knowledge, is a novel finding. These dose-response data may serve as useful contributions when assessing environmental risk with regard to marine oil pollution.