at SciVerse ScienceDirect

Environmental Pollution 174 (2013) 236e243

Contents lists available

Environmental Pollution

journal homepage: www.elsevier .com/locate/envpol

Distillation fraction-specific ecotoxicological evaluation of aparaffin-rich crude oil

Elisabeth Erlacher 1, Andreas P. Loibner, Romana Kendler, Kerstin E. Scherr*

University of Natural Resources and Life Sciences Vienna (BOKU), Department IFA-Tulln, Institute for Environmental Biotechnology, Konrad-Lorenz-Str. 20, 3430 Tulln, Austria

a r t i c l e i n f o

Article history:Received 4 June 2012Received in revised form13 November 2012Accepted 18 November 2012

Keywords:Risk assessmentCrude oilPetroleum hydrocarbonsHydrocarbon distillationSoil ecotoxicityEarthwormOstracodLuminescent bacteriaBioaccessibility

* Corresponding author.E-mail address: kerstin.scherr@boku.ac.at (K.E. Sch

1 Current address: RIFCON GmbH, Goldbeckstr. 13,

0269-7491/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.envpol.2012.11.031

a b s t r a c t

Crude oil is a complex mixture of petroleum hydrocarbons (PHC) with distinct chemical, physical and-toxicological properties relevant for contaminated site risk assessment. Ecotoxicological effects of crudeoil distillation fractions on luminescent bacteria (Vibrio fischeri), earthworms (Dendrobaena hortensis) andinvertebrates (Heterocypris incongruens) were tested using two spiked soils and their elutriates. Fraction2 (F2) had an equivalent carbon number (ECN) range of >10 to 16, and F3 from >16 to 39. F2 showeda substantially higher ecotoxicological effect than F3 for Vibrio and Dendrobaena. In contrast, severeinhibition of Heterocypris by the poorly soluble F3 is attributed to mechanical organ blockage. Immediatesequestration of PHC to the organic matter-rich soil effected reduced toxicity for all organisms. This studyindicates that a more differentiated consideration (i) of PHC mixtures based on ECN range and (ii) ofmodel soil properties employed for ecotoxicity testing should be included into PHC-contaminated siterisk assessment.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Risk assessment of sites contaminated with crude oil, petroleumhydrocarbons (TPH) or related products is based on the analyticaldetermination of the total TPH concentration, i.e. one single value,in most European countries. These are often non-risk basedapproaches, following the precautionary principle and are inaccordance with standard methods for petroleum hydrocarbonanalysis, e.g. ISO 16703:2004 or USEPA Method 8015B. Sincepetroleum hydrocarbons are highly complex mixtures ofcompounds (Wang et al., 1994) with significant differences in theirtoxicity (Blum and Speece, 1991), similar overall TPH concentra-tions with given bioaccessibility may not be similar in terms of therisk that they pose to a receptor. Thus, the determination of totalTPH concentrations does not necessarily deliver precise informa-tion on risks to human health and the environment. In order toprovide high safety standards, regulatory bodies define conserva-tive threshold values to account also for the more toxic petroleumhydrocarbon congeners that might be present at oil contaminated

err).69493 Hirschberg, Germany.

All rights reserved.

sites. However, in case of petroleum hydrocarbon contaminationassociated to low amounts of toxic and accessible compounds,the application of analytical threshold values is sometimesoverprotective.

Alternative approaches separate complex petroleum hydro-carbon mixtures into structural fractions (e.g. aliphatic andaromatic fraction) which represent groups of substances withcomparable structures or physico-chemical properties. Fraction-ation approaches include separation by e.g. solubility, boiling point,vapour pressure, carbon range, aromaticity and toxicity. One“fractionation approach” uses the concept of equivalent carbon (EC)numbers and was first established by the Total Petroleum Hydro-carbon Working Group (TPHCWG, 1997a,b, 1998a,b, 1999). Theequivalent carbon number (ECN) is an index for the relativeretention time on a gas chromatographic system normalised to n-alkanes (TPHCWG, 1997a,b), and allows for the arbitrary groupingof structurally different hydrocarbons that have approximatelysimilar chromatographic properties into user-defined EC numberranges. This fractionation by boiling point also reflects the distil-lation process; hydrocarbons within a certain chromatographicenvelope and of similar structure can be expected to be subject tophysicalechemical environmental processes, such as leaching andvolatilisation, to a similar extent (TPHCWG, 1997a). Depending onvariable amounts of different fractions in mineral oils and mineral

E. Erlacher et al. / Environmental Pollution 174 (2013) 236e243 237

oil products, different toxicities of hydrocarbon mixtures may beexpected at the same total petroleum hydrocarbon concentration.For example, a diesel, with an ECN range from 8 to 21, may be moretoxic than a heavy end distillate with an ECN range of 18 to >34(TPHCWG, 1998a,b) at the same total concentration measured insoil. The existing approaches, however, are mainly based on data ofenvironmental exposure models and human health data. Basicecotoxicity data are in most cases related to the water-solublecompound fractions but not to petroleum hydrocarbon contami-nated soil.

Ecotoxicity data on petroleum hydrocarbons can be found inpeer reviewed scientific literature. In most cases these data arerelated to petroleum hydrocarbon mixtures, petroleum products orindividual compounds of such mixtures. However, only limitedexperimental data on the ecotoxicity of specific fractions in soil isavailable (Cermak et al., 2010; Verbruggen, 2004). The presentwork aims to generate fraction-specific ecotoxicity data of differentparaffin-based crude oil distillation fractions via the application ofdifferent bioassays to spiked model soils.

In this study, fractions were directly achieved from an Austriancrude oil rich in paraffins. fractions F2, with a boiling point rangeequivalent to n-C10 to n-C16, and F3 (n-C16-n-C40) were obtained bycrude oil distillation under atmospheric pressure. Fraction F1 (<n-C10) is highly volatile and heavy losses have been observed duringspiking of soils earlier (Aichberger, 2005). Therefore, only F2 and F3were subjected to tests and analysis. There is evidence that soilswith different soil properties may alter the exhibition of toxiceffects due to varying sorption behaviour of the pollutants (Dornet al., 1998; Scheffer, 2002; Aichberger, 2005). In order to observevariations of effects, two dissimilar soils were tested. Soils werespiked with the individual distillation fractions and immediatelyapplied to ecotoxicity tests.

2. Material and methods

2.1. Crude oil distillation

Two fractions with a carbon number envelope equivalent to n-C11 to n-C16

(fraction 2, F2) and n-C16 to n-C40 (fraction 3, F3) were selected for distillation. Thedistillation of the paraffin-based oil originating from the Vienna Basin was per-formed in a commercially available distillation apparatus. F2 and F3 were obtainedunder ambient pressure at temperature ranges of 196e287 �C and 302 �C to themaximum steam temperature of 480 �C, respectively. The crude fractions werepurified by a clean-up step using a Vigreux column. The distillation procedure wasrepeated until a sufficient quantity of distillation fractions of at least 100 mL wasobtained. For reference, microbial conversion of this oil was studied in previousstudies (Scherr et al., 2007; Scherr et al., 2012).

2.2. Model soils

Two soils, labelled as H1 and AM, were used. The total organic carbon contents(TOC) were 5.3% for soil AM, and 0.1% for soil H1 (Table 1). Soils were artificiallycontaminated with each of the fractions for concentrations ranging from 100 to2000 mg kg�1 (fraction 2) and 500 to 10,000 mg kg�1 (fraction 3). The choice of the

Table 1Selected soil parameters of soils H1 and AM.

Soil property Soil

H1 AM

Particle size (d) distribution [% soil dry weight]Sand (d ¼ 2 mme63 mm) 82 8Silt (63e2 mm) 11 62Clay (<2 mm) 5 30Soil parametersSoil pH [�] 7.8 7.5WHCmax.

a 0.286 0.73TOCb 0.1 5.3

a Maximum water holding capacity; in g H2O/g soil dry weight.b Total Organic Carbon; in % (wt/wt) soil dry weight.

tested concentrations was based on previous experiments by Aichberger (2005) andErlacher (2004). For reference, soil AM was used in an earlier study on the sorptionof pentadecane (Ehlers et al., 2010), where it was designated Anmoor.

2.3. Ecotoxicity testing

2.3.1. Luminescence inhibition testThe LUMIStox� test (Dr. Bruno Lange GmbH, Obergrafendorf, Austria) was

applied to elutriates of the artificially contaminated soils. Soil elutriates wereprepared according to Hund and Traunspurger (1994). The assessment of adverseeffects to the bacteria Vibrio fischeri was accomplished by comparison of lightemission before and after addition of a contaminated elutriate to a bacterialsuspension and to an uncontaminated control (2% NaCl solution). Samples weretested in 5 dilution steps in triplicate. If testing was not possible at the same day ofelutriate preparation, elutriates were stored amaximum of 4 days at 4 �C in the dark.Test performance followed manufacturers’ standard operational procedure (SOP forLCK 480, 2000) and according to EN ISO 11348-3:1998.

The measurement of bioluminescence was performed using a Luminometer(LUMIStox LPG 259, Dr. Lange). The test was carried out at a constant temperature of15 �C in a cooling device (LUMIStherm LTG 053, Dr. Lange).

2.3.2. Ostracod mortality and growth inhibition testTest organisms were freshly hatched mussel shrimps of the species Heterocypris

incongruens (Crustacean) obtained from Ostracodtoxkit F� (MicroBioTests Inc.,Nazareth, Belgium). The test was originally developed for sediment testing but wasmodified for soil sample investigations according to Hirmann (2003). Test perfor-mance followed manufacturers’ standard operational procedure (MicroBioTests Inc.,2002) followed by an incubation period of 6 days at 25 �C.

2.3.3. Earthworm mortality testThe test was adopted fromOECD guideline 207 (OECD,1984). Earthworms of the

species Dendrobaena hortensis (Lumbricidae, Oligochaeta) were exposed to theartificially contaminated soils for a period of 14 days and mortality was determinedcompared to a control. The controls were uncontaminated samples of the same soilused in this experiment. Four replicates containing 500 g of soil (dry weight) wereprepared for each sample. Incubation time was 14 days at a room temperature of20 �C � 2 and at constant lighting. Uncontaminated soils H1 and AM served asnegative controls. Soil spiked with chloroacetamide in two concentrations (25 and40 mg kg�1) served as positive control.

The petroleum hydrocarbon content of spiked soils and soil elutriates weredetermined by GCeFID according to ÖNORM EN 14039 (2005); ISO 16703 (2004)and ÖNORM S2120 (1999), respectively, and as described by Scherr et al. (2007)and Aichberger et al. (2005).

2.3.4. Ultrasonic extraction of petroleum hydrocarbons from soilSoil samples were amended with 20 mL acetone p.a. and 10 mL RTW (Retention

Time Window) standard solution (n-decane [30 mL L�1] and n-tetracontane[30 mg L�1] in n-heptane, all from SigmaeAldrich, Vienna, Austria) followed by60 min ultrasound extraction. The solvent phase was separated and mixed withwater to remove acetone. Prior to filling into vials, the extracts were cleaned up byadding sodium sulphate and Florisil� (magnesium silicate; both from Sigma,Austria) in order to exclude water and polar substances from the analysis.

2.3.5. Ultrasonic extraction of petroleum hydrocarbons from soil elutriatesSoil elutriates were prepared in the same manner as elutriates used for the

LUMIstox� Test and in accordance with ISO 9377-2:2000. A soilewater suspensionin a ratio of 1:3.5 (w/w) was prepared. Each sample was measured in triplicate. Perreplicate, 200mg of spiked soil wasmixedwith 500mL of distilledwater and shakenon an end-over-end shaker (Beckman GS-6, Beckman Instruments GmbH, Munich,Germany) at 8 rpm for 24 h. Elutriates were centrifuged at 9500 rpm for 15 min ina Beckmann GH 3.8 centrifuge (Beckman Instruments GmbH, Munich, Germany),followed by decantation of the supernatant into a 500mL separating funnel where itsettled for a few hours. The elutriate was drained off into a Pyrex flask and theamount was measured on a weight basis. Each sample was amended with 10 mL ofRTW standard solution (n-decane [20 mL L�1] and n-tetracontane [20 mg L�1] in 1 Ln-heptane) and shaken at 10 rpm at an end-over-end shaker for 30 min. The solventphase was separated by using a separation funnel. The extract was cleaned-up usingNaSO4 and magnesium silicate as described above.

2.4. GCeFID analysis

The analytical procedure was adapted from ISO 16703 (2004). Extracts wereanalysed using an HP 5890 Series II, Hewlett Packard Gas Chromatograph (HewlettPackard, Palo Alto, California, USA) with an FID detector at 400 �C. A Hewlett PackardDB-5HTcapillarycolumn(length30m, ID0.25mm,film thickness0.1mm)wasused forseparation at a constant heliumflowof 1.80mL/min. The temperature programwas asfollows: 60 �C for 1min, followed by a 20 �C step perminute up to 380 �C, then held atthis temperature for 10 min. All samples were analysed in triplicate. TPH concentra-tions are, if not indicated differently, provided in mg per kg soil dry matter (DM).

E. Erlacher et al. / Environmental Pollution 174 (2013) 236e243238

2.5. Quantification of fractions

The analysis of the fractions C10eC16 and C16eC40 was basically in accordancewith ISO 16703 (2004) whereas the extraction, calibration and evaluation proce-dures were slightly modified. The area of the chromatogramwas integrated over then-C10 peak to the apex of the n-C16 peak and from the apex of the n-C16 peak to theapex of the n-C40 peak, respectively. An average response factor of the standarddiesel/lubricating oil calibration was used for quantification of each fraction. Themean recovery of each fraction from soil was 74.1% (F2, H1) and 75.2% (F3, H1) and102% for both fractions in the two soils. The overall mean recovery of TPH from soilranged from 86.4 to 119 %.

3. Results

3.1. Distillation of paraffin-based crude oil

Fraction 2 (F2) which corresponds to a boiling point range of196e287 �C was of clear liquid nature and showed yellow colour.The temperature of fraction 3 (F3) ranged from 302 to 483 �C. F3was a highly viscous distillate of dark brown to black colour.Purity of the fractions was determined by fractionated GCanalysis and subsequent calculation of impurities. Fig. 1 showsthe superimposed GCeFID chromatograms of single fractions anddemonstrates onlyminor overlapping of chromatograms of the twofractions. Fractions F3 and F3 had a purity grade of 88% and 96%after distillation, respectively (Table 2). Fig. 2 shows the visualappearance of the two different distillates.

3.2. Petroleum hydrocarbon analysis of spiked soils and elutriates

Petroleum hydrocarbon concentrations of the spiked soils andelutriates are shown in Tables 3 and 4, respectively. TPH values referto the total petroleum hydrocarbon concentration measured fromC10eC40. Concentrations of fractions were calculated from the totalmeasured value by adding the proportion of each fraction (87.8%and 96%, respectively; Table 2) to the total amount of petroleumhydrocarbons measured in the corresponding carbon range. Bysubtraction of the overlapping amount of each fraction (4% and10.1%, respectively, Table 2) from TPH values, the actual concen-trations of fractions were obtained.

For fraction 2, the results of the elutriate analysis of both soilsreflected the soil analysis data (Tables 3 and 4). Higher concentra-tions in soils resulted in higher concentrations in the correspondingelutriates (except for F2 in soil H1 where no elutriate data were

Fig. 1. GCeFID chromatograms of the two obtained fractions F2 (blue, lower retention time)of the references to colour in this figure legend, the reader is referred to the web version o

obtained). This was not observed in samples spiked with fraction 3.Elutriates of both soils exhibited no significant increase of F3 froma soil concentrations higher than 4500 mg kg�1 of soil dry matter.One reason for this observation may be the limited solubility ofpetroleum hydrocarbons of this fraction in water. The TPHCWG(1997a) estimated the water solubility of petroleum hydrocarbonfractions to be:

EC from > 10 to 12 : 3:4� 10�2 mg L�1

EC > 12� 16 : 7:6� 10�4 mg L�1

EC > 16� 21 : 2:5� 10�6 mg L�1

These estimations tremendously deviate from measured valuesin this study. However, the TPHCWG also reported such deviationsfrom other studies and attributes the differences to many factors,including differentmeasuring techniqueswith their inherent errorsin precision and accuracy, experimental conditions, the objectivesof experimental measurements, and the assumptions used inmaking estimates. The authors agree with this argument and pointout that it also has to be considered that the estimated values aboveare referred to n-alkanes only and are rough estimations. With themethod used in this study also iso-alkanes and cyclic and aromatichydrocarbons were measured which contribute to the sum ofmeasured petroleum hydrocarbons in a fraction. However, nocomprehensive conclusions can be drawn from the data ofTPHCWG and water solubility as the determining parameter on theobserved data cannot be clearly identified. Further, polarsubstances in the distillation products may be the cause of the lowsolubility of the fractions in water.

3.3. Ecotoxicity testing

3.3.1. LUMIstox� testRegarding the above discussed inconsistent results of the anal-

ysis, the establishment of a doseeresponse relationship betweenelutriate concentrations and bioluminescence inhibition data wasconsidered inappropriate. For soil H1, concentrations of fraction 2in the elutriate were below detection limit and for fraction 3 onlythree data points were available. In addition, a clear increase of the

and F3 (red, eluting later). Black circle indicates overlap of fractions. (For interpretationf this article.)

Table 3Petroleum hydrocarbon concentrations in soils: analysed TPH concentrations andcalculated fractions concentrations in mg kg�1. Standard deviations in parenthesis(n ¼ 3).

Petroleum hydrocarbon concentration in soil, in mg kg�1 soil dry matter

Soil H1 Soil AM

TPH(analysed)a

Fraction(calculated)

TPH(analysed)a

Fraction(calculated)

Fraction 294 (13) 67 (5) 129 (6) 100 (4)242 (35) 199 (18) 295 (7) 244 (2)405 (34) 339 (18) 597 (11) 521 (15)922 (67) 823 (30) 1.197 (34) 1.059 (33)1651 (42) 1476 (30) 2242 (53) 2050 (56)Fraction 3489 (24) 468 (24) 569 (63) 547 (61)1929 (245) 1848 (234) 2035 (189) 1953 (181)3205 (167) 3015 (174) 4508 (44) 4324 (37)4696 (58) 4500 (57) n.d. n.d.6330 (52) 6064 (47) 7430 (36) 7173 (39)n.d. n.d. 10,994 (30) 10,512 (30)

n.d. indicates no measurement at this level.a Between n-C10 and n-C40.

Table 2Calculated proportion of distillation fraction composition, referred to as fractionpurity.

Section of chromatogram,retention time windowbetween

Proportion of corresponding Fraction [%]

Fraction 2 Fraction 3

n-C10 and n-C16 87.8 4.0n-C16 and n-C40 10.1 96.0

E. Erlacher et al. / Environmental Pollution 174 (2013) 236e243 239

toxic effect was observed while on the other hand the concentra-tion in elutriates did not rise. These deviant observations and thelack of data gave rise to the decision to correlate soil concentrationsinstead of elutriate concentrations to toxic effects. Although thiscomparison does not reflect the true doseeresponse relationship, ithas the advantage to make results of the other ecotoxicity testscomparable to the LUMIstox� test.

Figs. 3 and 4 show the results of the LUMIstox� Test of the twopetroleum hydrocarbon fractions in soils AM and H1. They showthat bioluminescence inhibition caused by each of the fractionvaries significantly. In general, F2 in both tested soils led to highertoxicity values at the same concentration levels. Bioluminescenceinhibition of F3 samples did not significantly increase withconcentration in soil AM above 1953 mg kg�1. This was reflected ina way by the results of elutriate analysis: the concentration ofwater-soluble petroleum hydrocarbons did not increase froma concentration of 4324 mg kg�1 in soil, whereas bioluminescenceinhibition did not increase from a concentration of 1953 mg kg�1.The effect remained ranging from 23 to 30 % compared to thecontrols. As discussed above water solubility may play a major rolebut a comprehensive explanation still remains unresolved.

Variations were observed amongst the organic rich soil AM andthe sandy soil H1. Toxicity of both fractions was higher in soil H1.This is caused by different properties of the distillates and the soils,primarily sorption, hydrophobicity and solubility. Another factor isthe KOC value which is the “distribution coefficient” (Kd) normal-ized to total organic carbon content.

With decreasing water solubility and subsequent rising KOW, theKOC is increasing (Scheffer, 2002). Organic chemicals with high KOCvalues are strongly bound to the soil matrix. The observed varia-tions between toxicity in the two soils can therefore be attributed

Fig. 2. Visual appearance of fractions 2 and 3 obtained by distillation.

to an increased adsorption of the fractions, in particular F3, to theorganic matter content of soil AM.

3.3.2. Ostracod test3.3.2.1. Mortality. In soil H1 mortality caused by F2 ranged from13% up to 93% at the highest concentration of 1476 mg kg�1 (Fig. 5),whereas in soil AM mortality did not exceed 20% up to a concen-tration of 1794 mg kg�1 (Fig. 6). At the next and highest concen-tration level of soil AM (2.050 mg kg�1), a sudden increase to 71.7%of ostracod mortality was observed (Fig. 6).

By contrast, F3 in both soils exhibited higher mortality ratesthan F2 at similar tested concentrations. A mortality of 25% wasobserved at a concentration of 231 mg kg�1 for F3, whereas for F2only 10% of ostracods were affected at the same concentration level(Figs. 5 and 10). Comparing soils, significant lower toxic effectswere observed in soil AM. In soil H1 fraction 3 led to toxic effects atthe lowest applied concentration of 199 mg kg�1, whereas in soilAM a mortality rate higher than 20% was observed only at923 mg kg�1.

Table 4Petroleum hydrocarbon concentrations in elutriates: analysed TPH concentrationsand calculated fraction concentrations in mg L�1. Standard deviations in parenthesis(n ¼ 3).

Petroleum hydrocarbon concentration in soil elutriate [mg L�1 elutriate]

Soil H1 Soil AM

TPH(analysed)a

Fraction(calculated)

TPH(analysed)a

Fraction(calculated)

Fraction 20.1 (0.1) bdl 0.2 (0.0) bdl0.1 (0.0) bdl 0.3 (0.1) bdl0.8 (0.1) bdl 2.2 (0.5) 1.8 (0.4)1.4 (1.7) bdl 3.3 (1.0) 2.6 (0.9)1.6 (0.2) bdl 6 (0.4) 4.9 (0.3)Fraction 31.5 (0.4) 1.3 (0.3) 1.5 (0.3) 1.3 (0.2)4.9 (0.4) 4.4 (0.5) 3 (0.3) 2.6 (0.1)9.4 (1.5) 8.3 (1.4) 8.5 (1.0) 8 (0.9)9.4 (0.4) 8.5 (0.2) n.d. n.d.11 (1.2) 9.8 (1.1) 8.3 (3.0) 7.9 (2.9)n.d. n.d. 7 (1.5) 6.2 (1.5)

n.d. indicates no measurement at this level.bdl indicates calculated values are below detection limit of analytical method.

a Between n-C10 and n-C40.

Fig. 3. Results of the LUMIStox� test applied to soil H1 spiked with F2 (C) and F3 (-).Error bars represent standard error of mean (n ¼ 3).

Fig. 5. Results of the ostracod mortality test applied to soil H1 spiked with F2 (C) andF3 (-). Error bars represent standard error of mean (n ¼ 6).

E. Erlacher et al. / Environmental Pollution 174 (2013) 236e243240

3.3.2.2. Growth inhibition. Mean growth inhibition lower than 20%caused by fraction 2 was observed at a concentration range of 67e616 mg kg�1 in soil H1 and exceeded 50% up from a concentrationof 823 mg kg�1 (Fig. 7). In soil AM, growth inhibition was generallyslightly lower (Fig. 8).

Fraction 3, as observed in mortality testing, led to higher growthinhibition effects in both soils at similar concentration levels (Figs. 7and 8).

An explanation of the observed effects of fraction 3 to ostracodse which deviate significantly from the results of the earthwormmortality and the luminescent inhibition testemay be found in themode of action of the pollutant to these organisms. The toxicitymechanism that hydrophobic substances exert to microorganismcan be described as an interaction of the lipophilic pollutant withlipophilic regions of the cell membrane which further leads to anuptake of the substance and further to changes of structure andlosses of natural functions of the cell membrane (Sikkema et al.,1995). The toxic effect of hydrophobic substances to crustaceansis believed to follow other mechanisms, primarily mechanicalinterference rather than chemically induced toxicity (Percy, 1970;Percy and Mullin, 1977) and may lead to sublethal effects (Katz,1970; Moore and Dwyer, 1974; Rogerson et al., 1982). Ostracodstake up food from the surrounding water by producing a constantflow of food particles to their mouth. Fraction 3, which mainlycontains petroleum hydrocarbons of carbon numbers ranging from17 to 39, is of high viscosity. It is therefore assumed that theobserved adverse effects are sublethal effects and consequences ofreduced food uptake by clogging and/or blocking of feeding organs

Fig. 4. Results of the LUMIStox� test applied to soil AM spiked with F2 (C) and F3 (A).Error bars represent standard error of mean (n ¼ 3).

due to viscous hydrocarbons, and finally induced death caused bystarvation. These sublethal effects were also observed for othercrustaceans like amphipods (Busdosh, 1981) and lobsters (Atemaand Stein, 1974).

Another reason may be a negative effect on respiration. Respi-ration of ostracods is based on diffusion of dissolved oxygen inwater via the body surface and gill cover into the body. Gaseousexchange is promoted by movement of the limbs which producea water flow to the shell (Wade et al., 2004). It is thereforeconceivable that non-water soluble, highly viscous substances offraction 3 were transported to the gills and interfered with oxygen-exchange. As a consequence, the mussel shrimp dies of suffocation.Oil acting as a barrier to oxygen transfer between air and water wasalso observed by Kontagiannis and Barnett (1973). Bobra et al.(1989) and Hebert and Huhta (1984) reported that hereby theparticle size of dispersed oil particles inwatermay contribute to theextent of harmful effects in water.

Differences in soil properties were responsible for differingresults achieved for the two tested soils. As discussed above, a highcontent of organicmatter as well as higher quantities of clay and siltof soil AM create a higher sorption capacity which contributes to anincreased retention and sorption of hydrophobic substances. Asa consequence of reduced bioavailability, lower toxicity values wereobtained for soil AM (Figs. 6 and 8).

3.3.3. Earthworm mortality testThe earthworm mortality test showed similar results as the

LUMIstox� Test (Figs. 9 and 10). Fraction 2 in a concentration rangefrom 67 to 339 mg kg�1 caused no significant difference to thecontrol in soil H1 (�20%). At a concentration of 823 mg kg�1 the

Fig. 6. Results of the ostracod mortality test applied to soil AM spiked with F2 (C) andF3 (-). Error bars represent standard error of mean (n ¼ 6).

Fig. 7. Results of the ostracod growth inhibition test applied to soil H1 spiked with F2(C) and F3 (-). Error bars represent standard error of mean (n ¼ 6).

Fig. 8. Results of the ostracod growth inhibition test applied to soil AM spiked with F2(C) and F3 (-). Error bars represent standard error of mean (n ¼ 6).

Fig. 10. Results of the acute earthworm mortality test applied to soil AM spikedwith F2 (C) and F3 (-). Error bars represent standard error of mean (n ¼ 4).

Table 5�1 �1

E. Erlacher et al. / Environmental Pollution 174 (2013) 236e243 241

mortality rate ofDendrobaena hortensis increased to 100% (Fig. 9). Incontrast, toxicity of fraction 2 in soil AM was observed at concen-trations from 1059 to 2241 mg kg�1, but mortality did not exceed60%.

Fraction 3 led to significant lower mortality than fraction 2 inboth soils. Even the sandy soil H1 showed relatively low toxicity(<20%) up to a concentration of 4500 mg kg�1. At the next andhighest concentration level, mortality increased rapidly to 93%. Insoil AM mortality did not exceed 20% even at the highest concen-tration of 10,512 mg kg�1. An extrapolation of measured petroleumhydrocarbon concentrations and mortality showed that F3 would

Fig. 9. Results of the acute earthwormmortality test applied to soils H1 spiked with F2(C) and F3 (-). Error bars represent standard error of mean (n ¼ 4).

result in mortality of all earthworms at a sevenfold contaminationcompared to fraction 2.

Therefore the acute ecotoxicological potential of fraction 3 forearthworms may be classified as low.

Mean mortality in the negative control did not exceed 10%, thetest results were therefore accepted as valid. The positive controlsamples, soil spiked with chloroacetamide, were included in orderto ensure sufficient sensitivity of Dendrobaena hortensis. Accordingto ISO 11268-1:1993, LC50 values of this substance shall range from20 to 80mg kg�1. This test criterionwas fulfilled for soil H1whereasmean mortality in soil AM did not exceed 10%. It was therefore notpossible to calculate an LC50 value. Again soil properties of soil AMmay be responsible for those observed results. Since mortality didnot increase significantly over the test period, another soundexplanation, given by Yeardley et al. (1995), may be appropriate:a rapid degradation/decomposition of chloroacetamide duringincubation may have occurred.

The results of the earthwormmortality test showed clearly howsoil properties and physico-chemical properties of pollutantsinfluence adverse effects on soil organisms.

3.3.4. LCxx (lethal concentration) and ECxx (effective concentration)values

Table 5 shows the calculated EC50 and EC20 values of eachfraction and soil. Since fraction 3 led to bioluminescence inhibitionlower than 50%, a calculation of EC50 values for this fractionwas notfeasible. It has to be considered that these values refer to

Overview of calculated LC and EC values (in mg kg or mg L ) of all ecotoxicitytests.

TPH concentration in soil (mg kg�1 soil DM) or elutriate (mg L�1 elutriate)

Fraction 2 Fraction 3

H1 AM H1 AM

LUMIStox� testa

EC50 501 894 n.d. n.d.EC20 82 230 1290 2669Ostracod testb

LC50 1346 2225 774 1207LC20 985 2116 166 652EC50 1037 1955 868 971EC20 565 n.d. n.d. 205Earthworm mortality testb

LC50 1000 1853 5596 n.d.LC20 <800 776 4846 n.d.

n.d. indicates calculation was not feasible.a In mg TPH/L elutriate.b In mg TPH/kg of soil dry weight.

E. Erlacher et al. / Environmental Pollution 174 (2013) 236e243242

concentrations in soil only. A calculation of EC20 values referring toostracods growth inhibition could not be conducted for F2 in soilAM and F3 in soil H1. A precise calculation of LC50 values of acuteearthworm mortality was not feasible, therefore values wereroughly estimated.

4. Conclusions

The two distilled fractions showed for both the aquatic LUMI-stox� test and the terrestrial earthwormmortality test similar doseresponse relationships. For each, the short-chain hydrocarbonfraction 2 caused higher impacts to the organisms than the moreviscous fraction 3. By contrast, the ostracod test revealed resultswhich were not in accordance with these observations: Fraction 3was more toxic to mussel shrimp Heterocypris incongruens. Asdiscussed above, the adverse effects of highly viscous constituentsof F3 might be attributed to an impairment of respiration and/orfood uptake caused by adhering and clogging of organs which areinvolved in respiration and ingestion. It is recognised that short-chain hydrocarbons are more toxic to receptors than high molec-ular ones. Fritsche (1998) mentioned that short-chain petroleumhydrocarbons are, due to higher water solubility and further higherbioavailability, more toxic to receptors than petroleum hydrocar-bons of comparatively higher number of carbon atoms since vola-tility and water solubility are decreasing with increasing chainlength. Lower water solubility accounts for slower mass transferand decreased bioavailability which is intrinsically associated withtoxic effects of pollutants. In addition, a higher water solubility andbioavailability of these substances increases the toxic effect innatural environments (De Jonge et al., 1997).

Dorn et al. (1998) tested the acute toxicity of crude oilcontaminated soils on earthworms, plants and luminescentbacteria (Microtox� solid phase assay) and found the earthwormsto be more sensitive than bacteria. However, similar resultsregarding differences between soils with various organic contentswere obtained in their study.

Soils with a high content of organic matter show an increasedsorption performance and reduced bioavailability of organiccontaminants in soil (Loibner, 2003). This could also be observedfor soil AM in which both hydrocarbon fractions tend to result ina significantly lower ecotoxicological potential than in the sandysoil H1. Growth inhibition of the ostracod, on the other hand, didnot significantly differ between the soils. The differences foundbetween fractions in the two distinct soils may be attributed to theless hydrophobic character of fraction 2, which might lead toa minor adsorption to the sandy soil.

These results indicate that risks posed by crude oil contamina-tions are strongly dependent of the composition of petroleumhydrocarbon mixtures and the properties of the relevant fractions.Furthermore, soil matrix tremendously influences the extent ofadverse effects to soil organisms.

References

Aichberger, H., Hasinger, M., Braun, R., Loibner, A.P., 2005. Potential of preliminarytest methods to predict biodegradation performance of petroleum hydrocar-bons in soil. Biodegradation 16, 115e125.

Aichberger, K., 2005. Bioassays zur Ermittlung des Einflusses der Bioverfügbarkeitauf die Toxizität kohlenwasserstoff-kontaminierter Böden. PhD thesis, Univer-sity of Natural Resources and Life Sciences, Vienna, Austria.

Atema, J., Stein, L.S., 1974. Effects of crude oil on the feeding behaviour of the lobsterHomarus americanus. Environmental Pollution 6 (2), 77e86.

Bobra, A.M., Shin, W.S., Mackay, D., 1989. Acute toxicity of dispersed fresh andweathered crude oil and dispersants to Daphnia magna. Chemosphere 19 (8e9),1199e1222.

Blum, D.J.W., Speece, R.E., 1991. A database of chemical toxicity to environmentalbacteria and its use in interspecies comparisons and correlations. ResearchJournal of the Water Pollution Control Federation 63, 198e207.

Busdosh, M., 1981. Long-term effects of the water soluble fraction of Prudhoe baycrude oil on survival, movement and food search success of the arcticamphipod Boeckosimus (¼Onisimus) affinis. Marine Environmental Research5 (3), 167e180.

Cermak, J.H., Stephenson, G.L., Birkholz, D., Wang, Zhendi, Dixon, D.G., 2010.Toxicity of petroleum hydrocarbon distillates to soil organisms. EnvironmentalToxicology and Chemistry 29 (12), 2685e2694.

De Jonge, H., Freijer, J.I., Verstraten, J.M., Westerveld, J., Van der Wiele, F.W.M., 1997.Relation between bioavailability and fuel oil hydrocarbon composition incontaminated soils. Environmental Science and Technology 31, 771e775.

Dorn, P.B., Vipond, T.E., Salanitro, J.P., Wisniewski, H.L., 1998. Assessment of theacute toxicity of crude oils in soils using earthworms, microtox� and plants.Chemosphere 37 (5), 845e860.

EN ISO 11348-3, 1998. Wasserbeschaffenheit e Bestimmung der Hemmwirkung vonWasserproben auf die Lichtemission von Vibrio fischeri (Leuchtbakterientest) eTeil 3: Verfahren mit gefriergetrockneten Bakterien. International Organisationfor Standardisation, Geneva, Switzerland.

Ehlers, G.A.C., Forrester, S.T., Scherr, K.E., Loibner, A.P., Janik, L.J., 2010. Influence ofthe nature of soil organic matter on the sorption behaviour of pentadecane asdetermined by PLS analysis of mid-infrared DRIFT and solid-state 13C NMRspectra. Environmental Pollution 158, 285e291.

Erlacher, E., 2004. Etablierung und Optimierung eines Ökotoxizitätstests mit Den-drobaena hortensis (Lumbricidae; Oligochaeta) zur Bestimmung der Toxizitätrohölkontaminierter Böden. MSc thesis, University of Vienna, Vienna, Austria.

Fritsche, W. (Ed.), 1998. Environmental Microbiology. Principles and Applications.Spektrum Akademischer Verlag, Germany, ISBN 3-437-35008-0, p. 252.

Hebert, R., Huhta, V., 1984. Effects of modification of particle size of emulsions ofVenezuelan crude oil on feeding, survival and growth of marine zooplankton.Marine Environmental Research 4 (2), 121e134.

Hirmann, D., 2003. Anwendung von Bioassays für die Untersuchung von Altlastenkontaminiert mit polyzyklischen aromatischen Kohlenwasserstoffen (PAKs).PhD thesis, University of Natural Resources and Life Sciences, Vienna, Austria.

Hund, K., Traunspurger, W., 1994. Ecotox-evaluation strategy for soil bioremediationexemplified for a PAH-contaminated site. Chemosphere 29, 371e390.

ISO 9377-2, 2000. Water Quality e Determination of Hydrocarbon Oil IndexMethod Using Solvent Extraction and Gas Chromatography. InternationalOrganisation for Standardisation, Geneva, Switzerland.

ISO 11268-1, 1993. Soil Quality e Effects of Pollutants in Earthworms (Eisenia fetida).Part 1: Determination of Acute Toxicity Using Artificial Soil Substrate. Inter-national Organisation for Standardisation, Geneva, Switzerland.

ISO 16703, 2004. Soil Quality e Determination of Content of Hydrocarbon in theRange C10 to C40 by Gas Chromatography. International Organisation forStandardisation, Geneva, Switzerland.

Katz, L.M., 1970. The effects of water soluble fractions of crude oil on larvae of thedecapod crustacean Neopanope texana (Sayi). Environmental Pollution 5 (3),199e204.

Kontagiannis, J.E., Barnett, C.J., 1973. The effect of oil pollution on survival of thetidal pool copepod Tigriopus californicus. Environmental Pollution 4 (1), 69e79.

Loibner, A.P., 2003. Availability of Hydrophobic Organic Pollutants in Soil e Impli-cation on Environmental Receptors and Consequences for RemediationMeasures. Habilitation thesis, University of Natural Resources and Life SciencesVienna, Austria.

MicroBioTest Inc., 2002. OSTRACODTOXKITF� Chronic “Direct Contact Toxicity Testfor Freshwater Sediments” Bench Protocol. Standard Operational Procedure.

Moore, S.F., Dwyer, 1974. Effect of oil on marine organisms: a critical assessment ofpublished data. Water Research 8 (10), 819e827.

OECD Guidelines for the Testing of Chemicals, 1984. Earthworm, Acute Toxicity Test.Document 207. Organisation for Economic Co-operation and Development,Paris, France.

ÖNORM EN 14039, 2005. Charakterisierung von Abfällen e Bestimmung desGehalts an Kohlenwasserstoffen von C10 bis C40 mittels Gaschromatographie.Österreichisches Normungsinstitut/Austrian Institute for Standardisation,Vienna, Austria.

ÖNORM S2120, 1999. Bestimmung des Gesamtgehaltes an Kohlenwasserstoffen unddes Gehaltes an Kohlenwasserstoffen im Eluat im Abfällen mit IR-Spek-troskopie. Österreichisches Normungsinstitut/Austrian Institute for Stand-ardisation, Vienna, Austria.

Percy, J.A., 1970. Responses of arctic marine crustaceans to crude oil and oil-taintedfood. Environmental Pollution 10 (2), 155e162.

Percy, J.A., Mullin, T.C., 1977. Effects of crude oil on the locomotory activity of arcticmarine invertebrates. Marine Pollution Bulletin 8 (2), 35e40.

Rogerson, A., Berger, J., Grosso, C.M., 1982. Acute toxicity of ten crude oils on thesurvival of the rotifer Asplanchna sieboldi and sublethal effects on rates ofprey consumption and neonate production. Environmental Pollution 29 (3),179e187.

Scheffer, F., 2002. Lehrbuch der Bodenkunde/Scheffer/Schachtschabel, 15. Aufl.Spektrum Akademischer Verlag, Heidelberg, Germany.

Scherr, K., Aichberger, H., Braun, R., Loibner, A.P., 2007. Influence of soil fractions onmicrobial degradation behavior of mineral hydrocarbons. European Journal ofSoil Biology 43, 341e350.

Scherr, K.E., Lundaa, T., Klose, V., Bochmann, G., Loibner, A.P., 2012. Changes inbacterial communities from anaerobic digesters during petroleum hydrocarbondegradation. Journal of Biotechnology 157, 564e572.

Sikkema, J., De Bont, J.A.M., Poolman, B., 1995. Mechanisms of membrane toxicity ofhydrocarbons. Microbiological Reviews 59 (2), 201e222.

E. Erlacher et al. / Environmental Pollution 174 (2013) 236e243 243

SOP for LCK 480, 2000. Dr. Lange� LUMIstox Leuchtbakterientest. Leuchtbakter-ientest LCK 480 According to ISO 11348. Standard Operational Procedure,Düsseldorf, Germany.

TPHCWG, 1997a. Selection of Representative TPH Fractions Based on Fate andTransport Considerations. In: Total Petroleum Hydrocarbon Criteria WorkingGroup Series, vol. 3. Amherst Scientific Publishers, Amherst, Massachusetts.

TPHCWG, 1997b. Development of Fraction Specific Reference Doses (RfDs) andReference Concentrations (RfCs) for Total Petroleum Hydrocarbons. In: TotalPetroleum Hydrocarbon Criteria Working Group Series, vol. 4. Amherst Scien-tific Publishers, Amherst, Massachusetts.

TPHCWG, 1998a. Analysis of Petroleum Hydrocarbons in Environmental Media. In:Total Petroleum Hydrocarbon Criteria Working Group Series, vol. 1. AmherstScientific Publishers, Amherst, Massachusetts.

TPHCWG, 1998b. Composition of Petroleum Mixtures. In: Total Petroleum Hydro-carbon Criteria Working Group Series, vol. 2. Amherst Scientific Publishers,Amherst, Massachusetts.

TPHCWG, 1999. Human Health Risk-based Evaluation of Petroleum Release Sites:Implementing the Working Group Approach. In: Total Petroleum HydrocarbonCriteria Working Group Series, vol. 5. Amherst Scientific Publishers, Amherst,Massachusetts.

Verbruggen, E.M.J., 2004. Environmental Risk Limits for Mineral Oil (Total Petro-leum Hydrocarbons). RIVM Report 601501021/2004. National Institute forPublic Health and the Environment (RIVM), Bilthoven, NL.

Wade, S., Corbin, T., McDowell, L.-M., 2004. Critter Catalogue. A Guide to the AquaticInvertebrates of South Australian Inland Waters. Environmental ProtectionAuthority, Adelaide, Australia, ISBN 1 876562 67 6.

Wang, Z., Fingas, M., Li, K., 1994. Fractionation of a light crude oil and identificationand quantitation of aliphatic, aromatic, and biomarker compounds by GC-FIDand GC-MS, part II. Journal of Chromatographic Science 32, 367e382.

Yeardley Jr., R.B., Lazorchak, J.M., Pence, M.A., 1995. Evaluation of alternativereference toxicants for use in the earthworm toxicity test. EnvironmentalToxicology and Chemistry 14 (7), 1189e1194.