Toxicity and compositional analysis of aromatic hydrocarbon fractions of two pairs of undegraded and biodegraded crude oils from the Santa Maria (California) and Vienna basins

Organic Geochemistry Organic Geochemistry 37 (2006) 1885–1899 www.elsevier.com/locate/orggeochem

Toxicity and compositional analysis of aromatic hydrocarbon fractions of two pairs of undegraded and biodegraded crude oils from the Santa Maria (California) and Vienna basins Verena Reineke a, Ju¨rgen Rullko¨tter

a,*

, Emma L. Smith b, Steven J. Rowland

b

a

b

Institute of Chemistry and Biology of the Marine Environment (ICBM), Carl von Ossietzky University of Oldenburg, P.O. Box 2503, D-26111 Oldenburg, Germany Petroleum and Environmental Geochemistry Group, School of Earth, Ocean and Environmental Sciences, University of Plymouth, Drake Circus, Plymouth PL4 8AA, UK Available online 15 September 2006

Abstract The monoaromatic and total aromatic hydrocarbon fractions of two pairs of undegraded and moderately biodegraded crude oils from the Santa Maria basin (California) and the Vienna basin (Austria), all dominated by unresolved complex mixtures, were studied regarding their composition and toxicity towards the feeding rate of the marine mussel Mytilus edulis. Total aromatic and monoaromatic hydrocarbon fractions from sulphur-rich Monterey Formation crude oils were slightly more toxic than the fractions isolated from sulphur-lean Vienna basin oils. The ecotoxicity tests did not show any significant differences in toxicity of aromatic compounds from undegraded or in-reservoir biodegraded crude oils from the same oilfield although some differences in composition were observed. Organic sulphur compounds are suspected to cause the slightly higher toxicity of the aromatic hydrocarbon fractions from the Monterey oils.  2006 Elsevier Ltd. All rights reserved.

1. Introduction Oil spills are a severe threat of long-term damage to the environment and its ecosystem. Crude oils consist of a vast number of components with different ecotoxic potentials. Often low-molecular-weight compounds with their higher solubility in water and bioavailability are regarded as most problematic (e.g. Baussant et al., 2001). On the other hand, many of the less water-soluble lipophilic compounds preferentially accumulate in organisms. Some of these, e.g. the polycyclic aromatic hydrocarbons *

Corresponding author. E-mail address: [email protected] (J. Rullko¨tter).

(PAHs) with four and more rings, were proven to be highly toxic to animals and humans (Fent, 2003). A group of substances which has attracted less attention than the parent PAHs are alkylated aromatic compounds beyond methyl substitution. Many homologues and isomers of alkylaromatics occur in crude oils in low concentrations. Gas chromatograms of crude oils, especially of biodegraded oils, commonly show an unresolved complex mixture (UCM) of hydrocarbons with very similar physical properties and largely unknown structures (up to 250,000 substances as estimated by Sutton et al., 2005). Even in non-biodegraded crude oils UCMs may comprise 30% of total hydrocarbons (Revill, 1992). The UCM constituents appear to

0146-6380/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.orggeochem.2006.07.017

1886

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

be particularly resistant to biodegradation (Gough and Rowland, 1990) and may accumulate in the environment whenever an oil spill occurs. The aliphatic part of the UCM may essentially comprise ‘‘T-branched’’ alkanes and alkyl-substituted cyclic structures (Killops and Al-Juboori, 1990; Gough and Rowland, 1990; Gough et al., 1992; Warton et al., 1997). For the aromatic UCM, derivatives of indanes, tetralins, phenylbenzenes and alkylcyclohexylbenzenes were proposed as possible major constituents (Killops and Al-Juboori, 1990; Warton et al., 1999). Despite their relatively low solubility in water, hydrocarbons can easily accumulate in the fatty tissue of organisms, partly because cell permeability is high for hydrophobic substances. Alkanes with more than 11 carbon atoms are not acutely toxic to most organisms due to their low aqueous solubility and low chemical reactivity. Aromatic compounds, however, are known to be more problematic (Fent, 2003) due to their higher aqueous solubility and enhanced bioavailability. In a bioassay with the marine mussel Mytilus edulis (exposure time 24 h), Donkin et al. (1989, 1991) tested several individual hydrocarbons with respect to their potential to reduce mussel feeding rate, which is indicative for a non-specific mode of sublethal narcosis (Donkin et al., 1989). According to these studies, monoaromatics appear to be more toxic than substances with a higher number of aromatic rings (Donkin et al., 1989, 1991). Smith et al. (2001) found a reduction of mussel feeding rate for alkylated cyclohexyltetralins, which they had suggested as model compounds for monoaromatic UCMs. In the first study of the toxicity of an unresolved complex mixture of aromatic hydrocarbons, Rowland et al. (2001) revealed the toxicity of the monoaromatic UCM of a North Sea crude oil (partially degraded) in the same type of bioassay. The aim of the present study was to compare the monoaromatic and total aromatic hydrocarbon fractions of degraded and undegraded oils in order to test whether differences in composition influence toxicity. For this combination of structure analysis and toxicity tests, two pairs of undegraded and moderately biodegraded crude oils from two different provinces (Santa Maria basin, California, and Vienna basin, Austria) were chosen. Each pair of oils had identical source characteristics, and the Californian oils were moderately mature and sulphur-rich, whereas the others had a higher maturity and were lean in sulphur. The aspect of different sul-

phur contents was included in the study because Donkin et al. (1989) had found dibenzothiophene to be toxic to blue mussels. So far, there are only a few studies investigating the ecotoxic risk potential of sulphur-containing aromatic compounds (Donkin et al., 1989; Rhodes, 2002; in Colavecchia et al., 2004). 2. Materials and methods 2.1. Origin of sample materials The two Monterey Formation crude oils are from the Orcutt Field in the Santa Maria basin, California. The undegraded oil comes from API Well No. 04-083-02368 and the degraded oil from API Well No. 04-083-02322 (Tennyson and Isaacs, 2001). For the Vienna basin oils (Austria), the well locations have not been disclosed. North Sea blue mussels (Mytilus edulis) were collected from the rocky cove Port Quin in Cornwall, UK, for the experiments at the University of Plymouth and from the stony sand beach of Jade Bay near Wilhelmshaven, Germany, for the experiments in the Terramare Research Institute in Wilhelmshaven. 2.2. Preparation of aromatic hydrocarbon fractions from oil samples Approximately 500 mg whole oil were dissolved in 500 ll dichloromethane. A 20-fold excess of nhexane was added to ensure precipitation of asphaltenes. The n-hexane-soluble fraction was separated by medium-pressure liquid chromatography (Radke et al., 1980) into fractions of aliphatic/alicyclic hydrocarbons, aromatic hydrocarbons and polar heterocomponents (NSO). For further separation of the aromatic hydrocarbon fraction into mono-, di-, tri- and polyaromatic hydrocarbons, normal-phase HPLC was performed on a semipreparative Hypersil APS-2 column (250 mm · 10 mm · 10 lm) in series with a Hypersil APS-2 guard column (50 mm · 10 mm · 5 lm). The column oven for the main column was kept at 25 C. UV–vis detection with a diode array detector was performed at 254 nm for polyaromatics and 206 nm for monoaromatics. The retention times for the cuts between mono-, di-, tri- and polyaromatic hydrocarbon fractions were determined with the help of surrogate standard compounds and GC–MS analysis of 1 min

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

cut fractions of test separations with the aromatic hydrocarbon fraction of one of the Monterey oils. 2.3. Toxicological tests A widely-used sublethal assay was carried out, measuring the biological endpoint of reduction of feeding rate of the mussel, Mytilus edulis (Donkin et al., 1989; Smith, 2002). In short, toxicant solutions were prepared by adding solutions with different concentrations of the aromatic mixtures in 100 ll of acetone to 10 l of 50 lm-filtered seawater (15 C). Solutions were stirred for 2 h prior to use. Controls were carried out with acetone only. Groups of seven mussels (shell length 12–15 mm) were exposed to 1.4 l toxicant or control solution for 24 h in duplicates, ensuring gentle water movement with a Teflon-coated stirrer bar. The mussels were fed with an algal culture of Isochrysis galbana. For the determination of feeding rates, the mussels were transferred into individual beakers with fresh toxicant solution (200 ml) of the same exposure concentration. After an adaptation period of 30 min to make sure mussels were filtering again, a predetermined amount of algal culture was added to give a cell concentration of 30,000 cells/ml. An aliquot (20 ml) was immediately taken from each beaker and the cell count determined in triplicate per aliquot. A second aliquot was taken after 30 min, and the decline in cell concentration over 30 min was calculated. Cell counting was performed with a Beckman Coulter Z 2 (particles between 3 and 11 lm) or an industrial model D Coulter Counter (particles greater than 3 lm). For statistical analysis, StatGraphics Plus Version 5.1 (Manugistics Inc.) was used. Feeding rate data were tested for normality using standardised skewness and kurtosis. Afterwards, feeding rates were compared using analysis of variance (ANOVA) with differences at the <5% level considered significant (Smith, 2002). 2.4. Extraction of mussel tissue Mussels were dissected and the tissue was frozen. Alkaline digestion (NaOH, 20 min, 60 C) was followed by repeated extraction with n-hexane. For the removal of polar compounds from the mussel tissue, the total organic extracts were transferred to a Pasteur pipette filled with 4 cm silica gel 60 (particle size 40–63 lm, activated at 190 C for 2 h, then deactivated with 5% water). One hundred

1887

microliters of total organic extract were applied, the non-polar fraction was eluted with 4 ml of a mixture of n-hexane/dichloromethane (10:1, v/v) and analysed by gas chromatography (GC) and gas chromatography–mass spectrometry (GC–MS). 2.5. Analytical methods Whole oils, aliphatic hydrocarbon and aromatic hydrocarbon fractions were analysed for total sulphur contents by combustion in a LECO SC-444 instrument with a relative error of ±3%. The aromatic hydrocarbon fractions and the mussel tissue extracts were analysed on a Hewlett-Packard 5890 Series II gas chromatograph equipped with a Gerstel KAS 3 cold injection system and a fused silica column (J&W; 30 m, inner diameter 0.25 mm, coated with DB5, film thickness 0.25 lm). The injector temperature was programmed from 60 C (5 s) to 300 C (60 s) at 8 C/s. Helium was used as carrier gas, and the oven temperature was programmed from 60 C (1 min hold) to 305 C at a rate of 3 C/min, followed by an isothermal phase of 50 min. Quantitation of hydrocarbons in the mussel tissue was carried out via external calibration with dodecylbenzene. For compound and compound class identification, a similar GC system with a high-temperature column (J&W fused silica, 30 m, inner diameter 0.25 mm, coated with DB-5HT, film thickness 0.1 lm) was coupled to a Finnigan SSQ 710 B mass spectrometer, scanning the range of 50–650 u at 1 scan/s (ionisation energy 70 eV). The injector temperature was programmed from 60 C (5 s) to 350 C (120 s) at 10 C/s. The GC oven temperature was programmed as described above. 3. Results and discussion 3.1. Characteristics of the four crude oils The aromatic hydrocarbons comprise about 20% of the whole oils studied irrespective of their level of biodegradation (Table 1). According to the molecular maturity parameters, the Monterey Formation oils are less mature than the Vienna basin oils but have higher sulphur contents (Table 1). Both degraded oils have a biodegradation level of 4–5 after Peters and Moldowan (1993), as indicated, for example, by the pristane/n-C17 ratio. Biodegradation of organic sulphur compounds in crude oils has been less well studied than that of

1888

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

Table 1 Selected bulk compositional and maturity information for the Monterey (M) and Vienna basin oils (V); S = sulphur, sat = saturated hydrocarbons, ARO = total aromatic hydrocarbons, polar = NSO compounds and asphaltenes, MA = monoaromatic hydrocarbons, MPI-1 = 1.5 Æ (2-MP + 3-MP)/(P + 1-MP + 9-MP) Parameter

Undegraded Monterey oil

Degraded Monterey oil

Undegraded Vienna basin oil

Degraded Vienna basin oil

Depth (m) API gravity Sulphur content (%) % sat/aro/polar Pri/n-C17 ratio Pri/Phy C29 steranes 20S/(20S + 20R) C29 steranes 14b, 17b/(14a, 17a + 14b, 17b) MPI-1

1174b 22.7a 3.09 29/24/47 1.62 0.94 0.33 0.51 0.92

957b 22.3a 3.11 27/26/47 5.33 0.87 0.32 0.56 0.91

2000-3000c 29c 0.20 67/23/10 0.42 1.97 0.46 0.80 0.90

max. 1500c 22.6c 0.24 51/26/23 7.39 1.71 0.38 0.68 1.01

a b c

Rullko¨tter et al. (2001). Tennyson and Isaacs (2001). OMV (private communication); other data: this study.

hydrocarbons (see Kropp and Fedorak, 1998; for a review). It is known, however, that alkylated thiophenes can be degraded by oxidation of the alkyl chains (Fedorak et al., 1996), although alkylated condensed thiophenes are considered to be among the most recalcitrant compounds present in petroleum-contaminated environments (Bence et al., 1996). 3.2. Composition of aromatic hydrocarbon fractions Reconstructed total ion current (RIC) chromatograms from GC–MS analysis of the total aromatic (ARO) and monoaromatic (MA) hydrocarbon fractions of the four crude oils are shown in Fig. 1a–h. Labelled peaks are identified in Table 2. In the chromatograms of the monoaromatic hydrocarbon fractions several seemingly resolved peaks could not be assigned to single structures due to massive coelution. In general, the monoaromatic hydrocarbon fractions of the Monterey oils (Fig. 1a and c) contain more low-molecular-weight compounds than the total aromatic hydrocarbon fractions (Fig. 1d). This is not the case for the Vienna basin oils (Fig. 1e–h). Monoaromatic hydrocarbons in all oils are present up to high GC retention times, indicating the presence of high-molecular-weight monoaromatic compounds, i.e. highly substituted benzenoid structures. There are no obvious differences that can be attributed to biodegradation in the chromatograms of the Monterey oils. The number of resolved compounds in the degraded Vienna basin oil fractions (Fig. 1g and h) is smaller than in the undegraded counterparts (Fig. 1e and f).

3.2.1. Alkylated monoaromatic hydrocarbons Mass chromatography helped to reveal the distribution patterns of several compound classes in the monoaromatic hydrocarbon fractions (Fig. 2). The numbered compounds of each homologous series are identified in Table 2. The complexity of the distribution pattern within a compound class increases with the number of substituents on the aromatic nucleus as illustrated by the m/z 91 and 105 mass chromatograms (Fig. 2). The regular pattern of the monoalkylated benzenes (m/z 91) is clearly visible. There is only a relatively small UCM underlying the resolved peaks of the alkylbenzenes, which derives from more complex structures which either comprise a phenyl substituent (e.g. alkylbenzenes with a branched side chain) or produce an m/z 91 fragment ion in a more complex fragmentation. The alkyltoluene distribution pattern (m/z 105; meta before para and ortho; Ellis et al., 1992) with the possibility of three positional isomers at the benzene ring is already more complex. Comparison within the oil pairs shows that biodegradation has an impact on the composition of the alkylated benzenes and toluenes, since the long-chain (>C16) compounds were not found in the degraded oils. Table 3 illustrates some differences in monoaromatic hydrocarbon composition of the four oils. Alkylated indanes were found in all oils. Alkyl substituents of methylindanes (m/z 131) reach up to C17 with many isomers present (Fig. 2). Differences between the oils are only partly due to biodegradation slightly reducing the proportions of long-chain alkylindanes. Apparently, alkylindanes

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

1889

Fig. 1. Partial reconstructed total ion current chromatograms of monoaromatic (MA) and total aromatic (ARO) hydrocarbon fractions of the four oils used in this study.

1890

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

Table 2 Identification of labelled compounds in the chromatograms of the total and monoaromatic hydrocarbon fractions (Figs. 1 a-h and 2)

that a number of other compounds also produce this particular fragment ion.

Peak label

Compound

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Octylbenzene Tetradecyltoluene Methylnaphthalenes C2-naphthalenes C3-naphthalenes Dibenzothiophene Phenanthrene Methyldibenzothiophenes Methylphenanthrenes C2-dibenzothiophenes C2-phenanthrenes C3-phenanthrenes Squalene C2-indanes

3.2.2. Aromatic sulphur compounds All fractions of aliphatic/alicyclic hydrocarbons had sulphur contents below 0.05% (Table 1), indicating that non-aromatic organic sulphur compounds (OSC) do not elute in these fractions. Alkylthiolanes were found in the aromatic hydrocarbon fraction of the undegraded Monterey oil (Table 3). Alkylthiolanes are readily biodegraded and not commonly found in biodegraded oils (Fedorak et al., 1998), which explains their absence in the biodegraded Monterey oil. Several alkylated thiophenes elute in the monoaromatic hydrocarbon fraction, but there was no obvious series of homologues in the oil pair from the Monterey Formation. The monoaromatic hydrocarbon fractions of the sulphur-lean Vienna basin oils do not contain OSC. Only the total aromatic hydrocarbons of the Monterey oils contain alkylated benzothiophenes, of which the homologues with substituents up to

are quite resistant to biodegradation and thus can accumulate in the environment. The large UCM in the m/z 131 mass chromatogram (Fig. 2) suggests

Fig. 2. Partial reconstructed total ion current (RIC) and selected mass chromatograms of alkylbenzenes (m/z 91.1), alkyltoluenes (m/z 105.1) and alkylindanes (m/z 131.1) of the monoaromatic hydrocarbon fraction of the undegraded Monterey oil. For identification of labelled peaks see Table 2.

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

1891

Table 3 Systematics of selected compounds in the monoaromatic hydrocarbon fractions (MA) of the Monterey (M) and Vienna basin oils (V) MA

Undegraded M

Degraded M

Undegraded V

Degraded V

Alkylbenzenes (m/z 91.1)

n-Alkyl chain length from C6 to C25

n-Alkyl chain length from C6 to C13

n-Alkyl chain length from C7 to C26

Absent

Alkyltoluenes (m/z 105.1)

n-Alkyl chain length from C5 to C17

n-Alkyl chain length from C5 to C14, m- and p-toluenes start disappearing first

n-Alkyl chain length from C5 to C25

n-Alkyl chain length from C5 to C14, m- and p-toluenes start disappearing first

Alkylindanes (m/z 117.1)

Present in all oils, similar pattern for each oil pair, slight differences with biodegradation for long-chain n-alkylindanes

Alkylmethyl-indanes (m/z 131.1)

Similar pattern of peak doublets for both oils, slight differences with biodegradation

Similar pattern of peak doublets for both oils, slight differences with biodegradation

Alkylthiophene (m/z 97.1, 111.1, 125.1)

Several compounds

Several compounds

No alkylthiophenes found in MA fraction

Alkylthiolanes (m/z 101.1)

n-Alkyl chain length from C10 to C25, each has several isomers

n-Alkyl chain length from C10 to C19, each has several isomers

No alkylthiolanes found in MA fraction

o-most, p-least abundant

C6 were identified, higher homologues are masked by the UCM. Alkylated dibenzothiophenes are present in all oils, although only C0- to C3-dibenzothiophenes were detected in very low concentrations in the Vienna basin oils. 3.3. Toxicity of aromatic hydrocarbon mixtures After exposure to the toxicant solutions of the different aromatic hydrocarbon mixtures, mussels had accumulated some of these compounds in their tissues. Fig. 3 shows the unresolved complex mixture of total aromatic hydrocarbons of the undegraded Monterey Formation oil in the apolar fraction of a mussel tissue extract in comparison with the analysis of a control mussel. 3.3.1. Toxicity of aromatic hydrocarbons from the Monterey Formation oils The total aromatic hydrocarbon fractions of the two Monterey oils reduced mussel feeding rates by ca. 50% compared to control mussels as illustrated by the dose–response curves in Fig. 4. Body burdens of up to approximately 50 lg/g wet weight (300 lg/g dry weight) were accumulated by the majority of the mussels within 24 h. In the experiment with the degraded Monterey oil aromatic hydrocarbons, some mussel groups stored higher amounts of ca. 100 lg/g wet weight (600 lg/g dry weight) in their tissues, and one group of mussels even accumulated a body burden of 177 lg/g

m-most, p-least abundant

wet weight (1000 lg/g dry weight) from the undegraded Monterey Formation oil aromatic hydrocarbon fraction. However, these higher amounts did not lead to a further significant reduction of feeding rate. This is in accordance with data from Wraige (1997), who showed that gill tissue concentrations correlate better with the reduction of feeding rate than total body burdens, especially at high total body burden values. This is probably due to the fact that the gills (the presumed site of toxic action) have a certain capacity for lipophilic compounds which cannot be exceeded. Thus, additional hydrocarbons are accumulated in other parts of the mussel tissue where they do not have an effect on the feeding rate. Until saturation is reached, body burden analysis does correlate well with gill tissue concentration (Wraige, 1997). It is noteworthy that in this study, the saturation level was reached and exceeded after only 24 h of exposure. Due to the saturation effect, the two outliers were disregarded for the calculation of linear regressions. Usually, dose–response curves are sigmoidal and cover several orders of magnitude of the toxicant concentration. The steeply increasing part of a sigmoidal curve, disregarding the very low and high concentrations, can be approximated by a linear function. In this study, the concentrations applied essentially represent this almost linear part of the sigmoidal curve. Kooijman (1998) argued in favour of a linear function as a first approximation of tissue concentration, owing to the fact that basically each

1892

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

Fig. 3. Chromatograms of apolar fractions of the mussel tissue extracts exposed to acetone only (top) and aqueous solutions of the undegraded Monterey oil total aromatic hydrocarbon fraction at 500 lg/l seawater (bottom). See text for further explanations.

toxicant molecule contributes to the toxic effect to the same extent. The dose–response regression line for the undegraded Monterey oil experiment in Fig. 4 has a steeper slope than the regression line for the degraded oil. In addition, the regression coefficient is higher, indicating a better correlation between reduction of feeding rate and increasing tissue concentrations. Thus, there are indications that the total aromatic hydrocarbon fraction of the undegraded oil may be slightly more toxic than that of the degraded oil. The monoaromatic hydrocarbon fractions of the two Monterey oils reduced mussel feeding rates by 40% compared to control mussels, only one group of mussels had the feeding rate reduced by as much as 60%. Uptake of monoaromatic hydrocarbons resulted in body burdens of approximately 10–80 lg/g wet weight (60–480 lg/g dry weight). Unfortunately, some mussel tissue samples were contaminated during the clean-up procedure so that there is not enough evidence to conclude whether one of the two monoaromatic hydrocarbon fractions is more toxic than the other. 3.3.2. Toxicity of aromatic hydrocarbons from the Vienna basin oils The total aromatic hydrocarbon fractions of the Vienna basin oils did not cause a consistent reduc-

tion of feeding rate of the blue mussels (Fig. 5a), although the amounts of hydrocarbons taken up by the mussels (many up to 50 lg/g wet weight and several up to around 150 lg/g wet weight) were about as high as in the Monterey oil experiments. No mean value of feeding rate was different from the control mean, and regression calculation basically resulted in a line almost parallel to the x-axis (Fig. 5a). For both plots, the linear correlation coefficient was less than 0.01, indicating that there is no relation between concentrations of aromatic hydrocarbons and reduction of mussel feeding rate. The mussels exposed to the undegraded Vienna basin oil monoaromatic hydrocarbons accumulated only minor amounts of hydrocarbons and no reduction of feeding rate was observed (Fig. 5b, top). The mussels exposed to the corresponding fraction of the degraded Vienna basin oil, however, stored hydrocarbons in significant quantities, but surprisingly the feeding rate slightly increased instead of being reduced (Fig. 5b, bottom). This might imply that some monoaromatic compounds can stimulate mussel feeding rates at certain concentration levels. If this was the case, it could also explain the lack in toxic effect in the dose–response curve of the total aromatic fraction experiment (Fig. 5a) as an overlap of toxic effect and stimulation.

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

1893

Undegraded Monterey Formation oil total aromatic hydrocarbon fraction 0.4

120

0.3

( )

100 80

0.2

*

*

*

60

*

0.1

Feeding rate [%]

Feeding rate [L/h]

140

40

y = 0.23 - 0.0024 . *x r2 = 0.65 0.0 0

50

100

150

20 0 200

Total body burden [µg/g ww]

Degraded Monterey Formation oil total aromatic hydrocarbon fraction 0.4

120

( )

0.3

100 80

0.2

* *

*

60

Feeding rate [%]

Feeding rate [L/h]

140

40

0.1

y = 0.22 - 0.0007 . *x r2 = 0. 37 0.0 0

50

100

150

20 0 200

Total body burden [µg/g ww]

Fig. 4. Tissue concentration–response curves of the two total aromatic hydrocarbon fractions of the Monterey Formation oils. The control mean (triangle) and data points are shown with error bars (se) for feeding rate means (n = 7 or 6 if single outliers were present). Note that vertical error bars mainly comprise inhomogeneities in the behaviour of the living organisms, whereas analytical errors are much smaller and represented by the horizontal error bar of the control sample. Asterisks denote statistically significant differences from the control mean feeding rate. The tissue concentration of the control is shown ± standard deviation (n = 2). Data points in parentheses are outliers and were not used for calculation of linear regression. Regression equations and correlation coefficients are also given.

Stimulation of various biological endpoints has been reported before, although not for mussel feeding rate. Low-level contamination of oil in water can induce an increased immune response in other test organisms, leading to a stimulation of the observed parameter. Anderson (1977) cited several studies in his review which mention a stimulation of respiratory rate in fish at low concentrations, followed by a decline in respiration at higher concentrations of some oil mixtures. He also referred to stimulation of photosynthesis in phytoplankton at low-level oil exposure and subsequent suppression at higher levels. In general, toxicants sometimes stimulate reproduction at low concentrations, rather than reduce it (‘‘hormesis’’). The reason is largely unknown, but maybe the toxicant causes suppres-

sion of a secondary stress at low concentrations (Kooijman, 1998). It is possible that a short-term stimulation of feeding rate masks an underlying reduction of feeding rate but to our knowledge this has not been observed before. However, mixture toxicity in general (and enhancement and quenching of toxicity by mixtures in particular) is a relatively new realm of research and its complexity poses a lot of challenges (Grimme et al., 2000). 3.3.3. Comparison of the toxic effects caused by aromatic hydrocarbon mixtures The two sets of experiments were carried out in different laboratories, albeit by the same scientist. As a complicating factor, the mussels for the toxicity tests with the Monterey oil fractions were col-

1894

Undegraded Vienna basin oil monoaromatic hydrocarbon fraction b 0.4

Undegraded Vienna basin oil total aromatic hydrocarbon fraction 0.4 150

150

100

0.2

50

0.1

Feeding rate [L/h]

0.3

Feeding rate [%]

Feeding rate [L/h]

0.3

100

0.2

Feeding rate %]

a

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

50

0.1

y = 0.19 + 0.00001 * x r2 = 0. 0001 0.0

0.0

0

0

50 100 150 Total body burden [µg/g ww]

0

200

20 40 60 Total body burden [µg/g ww]

80

Degraded Vienna basin oil monoaromatic hydrocarbon fraction 0.4

Degraded Vienna basin oil total aromatic hydrocarbon fraction 0.4

150

150

50

0.1

100

0.2

50

0.1

y = 0.14 + 0.0034 * x r2 = 0.83

y = 0.18 + 0.00004 * x r2 = 0. 0043 0.0

0

0

50 100 150 Total body burden [µg/g ww]

Feeding rate [%]

Feeding rate [%]

100

0.2

Feeding rate [L/h]

0.3

0.3 Feeding rate [L/h]

0 100

200

0.0 0

20 40 60 Total body burden [µg/g ww]

80

0 100

Fig. 5. Tissue concentration–response curves of (a) the two total and (b) the two monoaromatic hydrocarbon fractions of the Vienna basin oils. The control mean (triangle) and data points are shown with error bars (se) for feeding rate means (n = 7 or 6 if single outliers were present). Asterisks denote statistically significant differences from the control mean feeding rate. The tissue concentration of the control is shown ±standard deviation (n = 7). Data points in parenthesis are outliers and were not used for calculation of linear regression. Regression equations and correlation coefficients are also given. See Fig. 4 caption for note on error bars.

lected in Cornwall, UK, whereas the Vienna basin oil experiments were carried out with mussels from the German North Sea coast. The latter mussels had lower control feeding rates which may indicate that their health was not optimal (Widdows et al., 1995). However, prior to the toxicity tests in Wilhelmshaven, random sampling did not show contamination of the mussels with hydrocarbons and neither did the tissue extracts of control mussels during the experiment. In order to be able to compare the two sets of data, experiments with several exposure concentrations of the total aromatic hydrocarbon fractions of the two biodegraded oils were repeated with blue mussels from Wilhelmshaven. Again, the control mussels had lower feeding rates than the ones from Cornwall and biological variance was higher. The reduction of feeding rates was essentially similar to

those shown in Fig. 4, bottom and Fig. 5a, bottom, i.e. the difference in response to oils of different origin was confirmed although the lower number of organisms used in this experiment caused some more scatter in the data. There are several groups of mussels from Wilhelmshaven in our exposure experiments which did not reduce their feeding rates in spite of high amounts of aromatic hydrocarbons in their tissue. It is possible that a previous exposure to toxic compounds in the natural environment of Jade Bay allowed a certain degree of adaptation and short-term tolerance to contaminated seawater. This process of adaptation has been observed before for, e.g., fish which were able to tolerate PAHs. It is possible that adaptation applies to a wide range of substances (‘‘cotolerance’’) (Fent, 2003).

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

3.4. Assessment of toxic effects caused by the aromatic hydrocarbons 3.4.1. Relative toxicities of total and monoaromatic hydrocarbon mixtures For comparing the toxicities of different compounds and mixtures, risk assessment and predictive toxicology have established tools such as the ECx value, which is the concentration required to induce a biological effect of x%. Often, EC50 values are used (Grimme et al., 2000). They can refer to water or tissue concentrations (WEC and TEC, respectively). The lower the concentration needed to induce a 50% inhibition of the biological endpoint in question, the more toxic is the compound or mixture. For the undegraded Monterey oil total aromatic hydrocarbon fraction, the calculated WEC50 value is 302 lg/l, obtained by plotting the feeding rates against the logarithmic nominal aqueous concentrations (r2 = 0.81). This is in the range of previously determined WEC50 values for monoaromatic and polyaromatic compounds (Table 4). Since narcosis is usually better related to tissue concentration, TEC50 values were calculated for the total and monoaromatic hydrocarbon fractions of the undegraded Monterey oil by plotting the feeding rates against the logarithmic tissue concentrations (total aromatic hydrocarbon fraction: r2 = 0.59; monoaromatic hydrocarbon fraction: r2 = 0.83). The TEC50 values of the total (119 lg/g wet weight) and monoaromatic hydrocarbon fracTable 4 WEC50 and TEC50 values of aromatic hydrocarbons affecting feeding rates of blue mussels Compound a

Toluene n-Propylbenzenea n-Pentylbenzeneb n-Heptylbenzeneb n-Octylbenzeneb 6-Cyclohexyltetralincd 7-Cyclohexyl-1-methyltetralinc,d 7-Cyclohexyl-1-propyltetralinc,d Biphenyla Naphthalenea Phenanthrenea Pyrenea Fluoranthenea a b c d

Donkin et al. (1989). Donkin et al. (1991). Wraige (1997). Smith (2002).

WEC50 (lg/l)

TEC50 (lg/g)

2347 862 123 93 79 24 42 62 295 922 148 >40 80

16 27 94 35 82 44 58 138 16 31 31 >189 627

1895

tions (116 lg/g wet weight) suggest that both hydrocarbon fractions are similarly toxic. The TEC50 values fit well with previously determined data of alkyl substituted monoaromatic compounds (Table 4). Individual TEC50 values of polyaromatic compounds range from 16 to 627 lg/g wet weight (Table 4). The monoaromatic hydrocarbons account for 55% of the undegraded Monterey oil total aromatic hydrocarbon fraction. If this percentage is used to find out how toxic the remaining aromatic compounds are, it turns out that all other, higher aromatic compounds combined give a TEC50 of 123 lg/g wet weight (calculation after Faust et al., 2003). However, these TEC50 values are based on 24 h exposures, inducing narcosis. Many of the aromatic components can cause additional toxic effects after longer exposure, even at very low concentrations. 3.4.2. The influence of exposure time on toxicity Exposure time influences the mechanisms of toxicity. In short-term experiments, the dominating toxic effect will be narcosis due to membrane dysfunction whereas in long-term experiments, specific mechanisms of toxic action will become more important (Hermens et al., 1984). This indicates that even if the whole toxic effect in our experiments was narcosis caused by the sum of aromatic hydrocarbons, the overall toxic effect after long-term exposure (e.g. after an oil spill) can be much more severe and must not be disregarded. The amounts of aromatic UCMs accumulated by the mussels in our 24 h tests are in the range of environmental concentrations of hydrocarbons in blue mussels affected by contamination in their natural habitats. For example, Smith (2002) found predominantly unresolved aromatic hydrocarbons at 392 lg/g dry weight in mussels from Whitby (UK), a level similar to the accumulated body burden of a monoaromatic UCM during exposure experiments (90 lg/g wet weight; 400 lg/g dry weight). Short et al. (2003) found that low levels of longterm crude oil pollution affect the sensitive life stages of marine organisms, especially fish embryos. Peterson et al. (2003) reported an unexpected persistence of toxic subsurface oil and chronic exposures at a sublethal level which still affects wildlife in the Alaskan coastal ecosystem near Prince William Sound where the Exxon Valdez grounded in March 1989. It was estimated that it will take 30 years after the spill to reach background levels again (Carls et al., 2001), during which time low-level contamination

1896

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

will continue to cause severe toxic effects. Although immediately after a spill one- and two-ring aromatic compounds are highly concentrated in the environment and cause toxic effects, the long-term toxicity is predominantly due to three-, four- and five-ring polyaromatic hydrocarbons, which were found to be washed out of oiled sediments continuously (Short et al., 2003; Peterson et al., 2003). In addition, one has to bear in mind that not only parent PAHs but also alkylated aromatic compounds can cause long-term damage (e.g. mutagenic effects; Thomas et al., 2002) and that they are present in higher abundance than the unsubstituted compounds. Thus, referring to several toxicity studies in marine environments, Rowland et al. (2001) criticised that toxic effects often were only ascribed to PAH contamination (sometimes with bad correlation coefficients) and the bulk of aromatic hydrocarbons in a UCM was disregarded. 3.4.3. Differences in toxicity due to the composition of the hydrocarbon mixtures Differences in the toxicity of oils have been reported previously (e.g. Rice et al., 1977), but this should be specified with respect to the type of toxic effect (narcotic vs. specific) and the extent of exposure (long- vs. short-term). Elevated toxicity is usually attributed to relatively high amounts of aromatic hydrocarbons in the oils and wateraccommodated fractions studied (e.g. Anderson, 1977; Rice et al., 1977; Betton, 1994). In general, unaltered crude oils contain more low-molecularweight aromatic compounds, whereas polyaromatic compounds are relatively more abundant in weathered and degraded oils. In all mixtures used in this study, the acutely toxic and very volatile compounds like C0–C4-benzenes were not present. However, as a function of biodegradation, the Monterey oils showed little variation, the Vienna basin oils none. This may (at least partly) be related to the way of preparation of the fractions used for exposure in which the volatile compounds were lost by evaporation. Consequently, differences in mussel response should be related to the composition of the mixtures governed by their differences in source. Since alkylbenzene and alkyltoluene distributions more strongly depend on biodegradation than on origin of the oils (Table 3), these compound classes appear not to account for the differences in the toxicity tests. The distribution of alkylindane isomers varies somewhat with origin of the oils (Table 3)

but since there are numerous representatives of this compound class in all the oils we consider it unlikely that the alkylindanes are responsible for the toxicity differences. The content and distribution of organic sulphur compounds strongly depend on the origin of the oils, and the more toxic Monterey oils are richer in these compounds than the Vienna basin oils. The only literature value for the toxic effect of organic sulphur compounds on blue mussels is the TEC50 value of 14 lg/g wet weight for dibenzothiophene (Donkin et al., 1989). Fig. 6 shows a summed ion chromatogram of C1- to C3-dibenzothiophenes (m/z {198 + 212 + 226} extracted from the tissues of mussels exposed to total aromatic hydrocarbon fractions from the undegraded Monterey Formation (top) and Vienna basin oil (bottom). The relative amount of sulphur compounds to pure hydrocarbons in the tissue is much higher after exposure of the mussels to the former oil. The monoaromatic UCM from a Gullfaks oil used in a previous toxicity test with the approximate WEC50 of 520 lg/l and TEC50 of 500 lg/g (Smith, 2002) was investigated for organic sulphur compounds but none were found after GC–MS analysis. The typical sulphur content of Gullfaks oil is <1%. The lower toxicity determined for the Gullfaks monoaromatic UCM compared to the Monterey Formation monoaromatic fraction may therefore eventually be explained by the absence or very low abundance of OSC in the former oil. The Vienna basin oil fractions may lie in the same range of toxicity as the Gullfaks UCM if applied to mussels of equal health and sensitivity. 4. Conclusions Monoaromatic and total aromatic hydrocarbon fractions with significant UCMs of oils from two different provinces (Santa Maria basin, CA, Monterey Formation, and Vienna basin) were toxic to blue mussels (Mytilus edulis). The Monterey oil fractions induced a slightly higher toxic effect on blue mussels than the Vienna basin oil fractions. Differences in toxicity between undegraded and degraded crude oils from the Monterey Formation were small, and the two Vienna basin oils did not differ in toxicity at all. Monoaromatic compounds have a higher narcotic toxicity during short-term exposure and contribute chiefly to the toxicity of the total aromatic hydrocarbon mixture due to this higher toxicity and their high relative amount (55% of total aromatic hydrocarbon fractions).

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

1897

Fig. 6. Partial chromatograms of reconstructed total ion current and summed mass chromatograms of C1- to C3-dibenzothiophenes (m/z [198.2 + 212.2 + 226.2]) in the mussel tissue after exposure to the total aromatic hydrocarbon fractions at the (nominal) concentration of 500 lg/l seawater of the undegraded Monterey oil (top) and the undegraded Vienna basin oil (bottom).

1898

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899

Alkylbenzene and alkyltoluene distributions in all oils were fairly uniform. The observed differences in indane- and tetralin-derived aromatic hydrocarbons appear to be too small to carry a significant responsibility for the difference in toxicity between the two oil pairs. Since the main source difference between the Monterey and Vienna basin oils is the high sulphur content of the former, aromatic sulphur compounds may be another suspect adding to the joint effect of the aromatic hydrocarbons and to induce the slightly higher toxic effect of the Monterey oils, either as more potent narcotics or via a specific mechanism of toxic action. A higher relative abundance of alkylated dibenzothiophenes extracted from the tissue of mussels exposed to the Monterey oils seems to support this. Further support of this preliminary conclusion is required, however. Since the availability of conventional crude oil will cease one day but oil will still be needed as an important energy source, more ‘‘low-quality oils’’ may be produced. These are often characterised by high sulphur contents. Thus, the high toxicity of organic sulphur compounds is highly relevant and action should be taken to prevent oils and especially sulphur-rich oils from entering ecosystems, particularly if the higher toxicity of the organic sulphur compounds becomes substantiated. Acknowledgements We thank Dr. Paul Lillis (US Geological Survey, Denver) and OMV (Vienna) for providing Monterey Formation and Vienna basin oil samples, respectively. For helpful discussions, we thank Prof. Simon C. Brassell (University of Indiana) and Dr. Paul Sutton (University of Plymouth). In addition, we thank Prof. Gerd Liebezeit (Terramare Research Centre, Wilhelmshaven), Prof. Peter Jaros, Prof. Peter Ko¨ll and Dr. Gerd-Peter Zauke (University of Oldenburg) for providing space in their laboratories and/or providing instruments. For analytical support, we like to thank Bernd Kopke, Kerstin Adolph, Andreas Lu¨k, Ina Glemnitz, Mirja Bardenhagen, Hanna Brengelmann, Edith Kieselhorst (University of Oldenburg) and Martin Canty (University of Plymouth). Critical reviews by Prof. Zeev Aizenshtat and an anonymous referee are highly appreciated. Guest Associate Editor—J. Curiale

References Anderson, J.W., 1977. Responses to sublethal levels of petroleum hydrocarbons: are they sensitive indicators and do they correlate with tissue contamination? In: Wolfe, D.A. (Ed.), Fate and Effects of Petroleum Hydrocarbons in Marine Ecosystems and Organisms. Pergamon Press, Oxford, pp. 95– 114. Baussant, T., Sanni, S., Jonsson, G., Skadsheim, A., Boerseth, J.F., 2001. Bioaccumulation of polycyclic aromatic compounds: 1. Bioconcentration in two marine species and in semipermeable membrane devices during chronic exposure to dispersed crude oil. Environmental Toxicology and Chemistry 20, 1175–1184. Bence, A.E., Kvenvolden, K.A., Kennicutt II, M.C., 1996. Organic geochemistry applied to environmental assessments of Prince William Sound, Alaska, after the Exxon Valdez oil spill – a review. Organic Geochemistry 24, 7–42. Betton, C.I., 1994. Oils and hydrocarbons. In: Calow, P. (Ed.), Handbook of Ecotoxicology, vol. 2. Blackwell Scientific, Oxford, pp. 244–263. Carls, M.G., Babcock, M.M., Harris, P.M., Irvine, G.V., Cusick, J.A., Rice, S.D., 2001. Persistence of oiling in mussel beds after the Exxon Valdez oil spill. Marine Environmental Research 51, 167–190. Colavecchia, M.V., Backus, S.M., Hodson, P.V., Parrott, J.L., 2004. Toxicity of oil sands to early life stages of fathead minnows (Pimephales promelas). Environmental Toxicology and Chemistry 23, 1709–1718. Donkin, P., Widdows, J., Evans, S.V., Worrall, C.M., Carr, M., 1989. Quantitative structure–activity relationships for the effect of hydrophobic organic chemicals on rate of feeding by mussels (Mytilus edulis). Aquatic Toxicology 14, 277–294. Donkin, P., Widdows, J., Evans, S.V., Brinsley, M.D., 1991. QSARs for the sublethal responses or marine mussels (Mytilus edulis). The Science of the Total Environment (109–110), 461–476. Ellis, L., Kagi, R.I., Alexander, R., 1992. Separation of petroleum hydrocarbons using dealuminated mordenite molecular sieve. I. Monoaromatic hydrocarbons. Organic Geochemistry 18 (5), 587–593. Faust, M., Altenburger, R., Backhaus, T., Blanck, H., Bo¨decker, W., Gramatica, P., Hamer, V., Scholze, M., Vighi, M., Grimme, L.H., 2003. Joint algal toxicity of 16 dissimilarly acting chemicals is predictable by the concept of independent action. Aquatic Toxicology 63, 43–63. Fedorak, P.M., Coy, D.L., Peakman, T.M., 1996. Microbial metabolism of some 2,5-substituted thiophenes. Biodegradation 7, 313–327. Fedorak, P.M., Payzant, J.D., Montgomery, D.S., Westlake, D.W., 1998. Microbial degradation of n-alkyl tetrahydrothiophenes found in petroleum. Applied and Environmental Microbiology 54, 1243–1248. ¨ kotoxikologie: Umweltchemie, Toxikologie, Fent, K., 2003. O ¨ kologie. Thieme, Stuttgart, 332pp. O Gough, M.A., Rowland, S.J., 1990. Characterization of unresolved complex mixtures of hydrocarbons in petroleum. Nature 344, 648–650. Gough, M.A., Rhead, M.M., Rowland, S.J., 1992. Biodegradation studies of unresolved complex mixtures of hydrocarbons: model UCM hydrocarbons and the aliphatic UCM. Organic Geochemistry 18, 17–22.

V. Reineke et al. / Organic Geochemistry 37 (2006) 1885–1899 Grimme, L.H., Altenburger, R., Backhaus, T., Faust, M., Bo¨decker, W., Scholze, M., 2000. Kombinationswirkungen ¨ kotoxikologie. Zeitschrift von Umweltchemikalien in der O ¨ kotoxikologie 12 (4), 226–234. fu¨r Umweltchemie und O Hermens, J., Canton, H., Janssen, P., de Jong, R., 1984. Quantitative structure–activity relationships and toxicity studies of mixtures of chemicals with anaesthetic potency: acute lethal and sublethal toxicity to Daphnia magna. Aquatic Toxicology 5, 143–154. Killops, S.D., Al-Juboori, M.A.H.A., 1990. Characterisation of the unresolved complex mixture (UCM) in the gas chromatograms of biodegraded petroleums. Organic Geochemistry 15, 147–160. Kooijman, S.A.L.M., 1998. Process-oriented descriptions of toxic effects. In: Schu¨u¨rmann, G. (Ed.), Ecotoxicology. Wiley & Spektrum, New York, Heidelberg, pp. 484–520. Kropp, K.G., Fedorak, P.M., 1998. A review of the occurrence, toxicity, and biodegradation of condensed thiophenes found in petroleum. Canadian Journal of Microbiology 44, 605–622. Peters, K.E., Moldowan, J.M., 1993. The Biomarker GuideInterpreting Molecular Fossils in Petroleum and Ancient Sediments. Prentice Hall, Englewood Cliffs. Peterson, C.H., Rice, S.D., Short, J.W., Esler, D., Bodkin, J.L., Ballachey, B.E., Irons, D.B., 2003. Long-term ecosystem response to the Exxon Valdez oil spill. Science 302, 2082–2086. Radke, M., Willsch, H., Welte, D.H., 1980. Preparative hydrocarbon group type determination by automated medium liquid chromatography. Analytical Chemistry 52 (3), 406–411. Revill, A.T., 1992. Characterisation of unresolved complex mixtures of hydrocarbons by degradative methods. PhD thesis, University of Plymouth. Rhodes, S.M., 2002. Investigating the effects of dimethylated PAH on the embryo/larval development of the Japanese medaka (Oryzias latipes). MSc thesis. University of Waterloo, Waterloo, ON, Canada. Rice, S.D., Short, J.W., Karinen, J.F., 1977. Comparative oil toxicity and comparative animal sensitivity. In: Wolfe, D.A. (Ed.), Fate and Effects of Petroleum Hydrocarbons in Marine Ecosystems and Organisms. Pergamon Press, Oxford, pp. 78– 94. Rowland, S.J., Donkin, P., Smith, E.L., Wraige, E., 2001. Aromatic hydrocarbon ‘‘humps’’ in the marine environment: unrecognized toxicants? Environmental Science & Technology 35, 2640–2644. Rullko¨tter, J., Isaacs, C.M., Party, C.S., 2001. Compilation of geochemical data and analytical methods: cooperative mon-

1899

terey organic geochemistry study. In: Isaacs, C.M., Rullko¨tter, J. (Eds.), The Monterey Formation: From Rocks to Molecules. Columbia University Press, New York Chichester, West Sussex, pp. 399–460. Short, J.W., Rice, S.D., Heintz, R.A., Carls, M.G., Moles, A., 2003. Long-term effects of crude oil on developing fish: lessons from the Exxon Valdez oil spill. Energy Sources 25, 509–517. Smith, E.L., 2002. Unresolved complex mixtures of aromatic hydrocarbons in the marine environment: toxicity, solubility and photodegradation studies. PhD thesis, University of Plymouth. Smith, E.L., Wraige, E., Donkin, P., Rowland, S.J., 2001. Hydrocarbon humps in the marine environment: synthesis, toxicity, and aqueous solubility of monoaromatic compounds. Environmental Toxicology and Chemistry 20 (11), 2428–2432. Sutton, P.A., Lewis, C.A., Rowland, S.J., 2005. Isolation of individual hydrocarbons from the unresolved complex hydrocarbon mixture of a biodegraded crude oil using preparative capillary gas chromatography. Organic Geochemistry 36, 963–970. Tennyson, M.E., Isaacs, C.M., 2001. Geologic setting and petroleum geology of Santa Maria and Santa Barbara Basins, Coastal California. In: Isaacs, C.M. (Ed.), The Monterey Formation – From Rocks to Molecules. Columbia University Press, New York Chichester, West Sussex, pp. 206–229. Thomas, K., Balaam, J., Barnard, N., Dyer, R., Jones, C., Lavender, J., McHugh, M., 2002. Characterisation of potentially genotoxic compounds in sediments collected from United Kingdom estuaries. Chemosphere 49, 247–258. Warton, B., Alexander, R., Kagi, R.I., 1997. Identification of some single branched alkanes in crude oils. Organic Geochemistry 27 (7–8), 465–476. Warton, B., Alexander, R., Kagi, R.I., 1999. Characterisation of the ruthenium tetroxide oxidation products from the aromatic unresolved complex mixture of a biodegraded crude oil. Organic Geochemistry 30, 1255–1272. Widdows, J., Donkin, P., Brinsley, M.D., Evans, S.V., Salkeld, P.N., Franklin, A., Law, R.J., Waldock, M.J., 1995. Scope for growth and contaminant levels in North Sea mussels Mytilus edulis. Marine Ecology Progress Series 127, 131–148. Wraige, E., 1997. Studies of synthesis, environmental occurrence and toxicity of unresolved complex mixtures (UCMs) of hydrocarbons. PhD thesis, University of Plymouth.