2992

Environmental Toxicology and Chemistry, Vol. 24, No. 11, pp. 29922999, 2005
2005 SETAC
Printed in the USA
0730-7268/05 $12.00 .00
PREDICTING SINGLE AND MIXTURE TOXICITY OF PETROGENIC POLYCYCLIC
AROMATIC HYDROCARBONS TO THE COPEPOD OITHONA DAVISAE
CARLOS BARATA,* ALBERT CALBET, ENRIC SAIZ, LAURA ORTIZ, and JOSEP MARIA BAYONA
Laboratory of Environmental Toxicology, Universitat Politecnica de Catalunya, CN 150 Km 14.5, Terrassa 08220, Spain
Institut de Cie`ncies del Mar-CMIMA, Consejo Superior de Investigaciones Cient cas. P. Mar tim de la Barceloneta 37-49,
Barcelona 08003, Spain
Department of Environmental Chemistry, Institut dInvestigacions Qu miques i Ambientals de Barcelona, Consejo Superior de
Investigaciones Cient cas, Jordi Girona 18, Barcelona 08034, Spain
(Received 15 March 2005; Accepted 19 May 2005)
AbstractIn the present study, the acute toxicity of 10 polycyclic aromatic hydrocarbons (PAHs) associated with the Prestige
fuel oil spill (Spain, 2002) were evaluated, either as single substances or in mixtures, in adults of the copepod Oithona davisae.
All but dimethylphenanthrene had negative effects on O. davisae survival at concentrations below their water solubility, with 48-
h median lethal concentrations for naphthalene and pyrene of 56.1 and 0.8 mol/L, respectively, making these the least and most
toxic compounds. Polycyclic aromatic hydrocarbons had narcotic effects on copepods, as evidenced by the lack of motility at lower
concentrations than those causing death. Naphthalene showed the greatest narcotic effects, and phenanthrene showed minor effects.
Acute toxicity of the tested PAHs was inversely related (r
2
0.9) with their octanolwater partition coefcient, thereby conrming
the validity of the baseline quantitative structureactivity regression models for predicting the toxicity of PAH compounds in
copepod species. When supplied in mixtures, the toxic effect of PAHs was additive. These results indicate that the many PAHs in
an oil spill can be considered unambiguous baseline toxicants (class 1) acting additively as nonpolar narcotics in copepods; hence,
their individual and combined toxicity can be predicted using their octanolwater partition coefcient.
KeywordsMixture toxicity Polycyclic aromatic hydrocarbons Zooplankton Copepod
INTRODUCTION
Marine coastal organisms are frequently exposed to a range
of hydrocarbon concentrations as a result of oil spills, oil seep-
age, chronic coastal oil pollution from harbors, and to a small
extent, biogenic hydrocarbon production [1,2]. Marine zoo-
plankton, particularly copepods, are an important connection
between planktonic primary production and upper-trophic-lev-
el harvestable yield [3]. Given the importance of this link in
the marine food chain, it is necessary to establish the envi-
ronmental hazards of petroleum hydrocarbons on copepod spe-
cies [1,4].
The persistent toxicity of petroleum hydrocarbons largely
results from polycyclic aromatic hydrocarbons (PAHs), be-
cause other toxic components, including benzenes, toluenes,
and xylenes, are volatile and have shorter residence times in
the aquatic environment [5]. Petrogenic PAHs typically occur
in complex mixtures [2,6] and are considered to be a homo-
geneous group that exerts its toxicity by interfering with mem-
brane uidity, a phenomenon that is termed nonpolar narcosis
[7]. A theoretical framework has been established to predict
the toxicity of nonpolar narcotic PAH mixtures in aquatic or-
ganisms using quantitative structureactivity relationship
(QSAR) models and the concentration addition concept for
predicting mixture toxicity [813]. Quantitative structureac-
tivity relationships are mathematical models that relate quan-
titative measures of chemical structure to biological effects
and, hence, can be used for predicting the potential hazards
of untested compounds [14]. Experimental evidence from dif-
ferent species has shown that the single-substance toxicity of
* To whom correspondence may be addressed
(barata@intexter.upc.edu).
many PAHs can be predicted from their lipophilic properties
(i.e., octanolwater partition coefcient [K
ow
]) and that in mix-
tures, they act additively to produce toxic responses [8
10,12,13]. Of particular interest are those models derived for
aquatic amphipods that used QSARs to estimate median lethal
concentration (LC50) effects and the concentration addition
concept to predict the toxicity of complex mixtures [811].
The use of a sum PAH model [8] resulted in improved pre-
diction of observed toxicity for PAH-contaminated sediments.
Nevertheless, these empirical models have been shown to vary
across species, compounds, and environmental factors, such
as ultraviolet (UV) radiation that photoactivates many PAH
compounds [9,11], thus suggesting that PAH additive-toxicity
models should be derived and validated for particular case
studies.
In this sense, most copepod species show vertical circadian
migrations, feeding in surface waters during the night and
migrating to deeper waters during daylight [15]. Therefore,
effects of UV radiation on PAH toxicity are considered to be
negligible to these organisms.
Petrogenic PAHs in the water column comprise low-mo-
lecular-weight hydrocarbons dominated by compounds such
as naphthalene, phenanthrene, and their alkylated derivatives
[2,6]. Thus, laboratory-derived risk assessment procedures of
PAHs to planktonic marine communities should be focused on
the toxicity of parental and alkylated low-molecular-weight
hydrocarbons in the absence of, or at low levels of, UV ra-
diation. Despite the existence of substantial information re-
garding single-substance and mixture toxicity to aquatic or-
ganisms, information concerning marine copepods is restricted
to that of single compounds in meiobenthic species [16,17],
with only few studies having addressed PAH toxicity to plank-
Single and mixture toxicity of PAHs to Oithona davisae Environ. Toxicol. Chem. 24, 2005 2993
tonic species. Studies conducted with the estuarine copepod
Eurytemora afnis indicated that naphthalene toxicity increas-
es proportionally with the number of methyl substitutions [18]
and that low-molecular-weight PAHs induce narcosis, as ev-
idenced by lack of motility at sublethal concentrations [1, 18,
19]. Therefore, the above-mentioned toxicity features should
be taken into account when addressing the study of single-
substance and mixture toxicity of PAHs.
The aim of the present study was to use the theoretical
framework of nonpolar narcosis and additive toxicity of PAH
mixtures to predict the toxicity of PAHs associated with the
fuel oil of the Prestige spill in the copepod Oithona davisae.
In particular, we tested whether acute responses to single and
mixture combinations of petrogenic PAHs could be predicted
from QSAR models based on their K
ow
and by using the con-
centration addition mixture-toxicity approach [8].
The oil tanker Prestige sank off Galicia (northwest Spain)
in the year 2002, spilling 63,000 tons of number 2 fuel oil
into the open sea [20]. Number 2 oil is a middle distillate
normally used for home heating and diesel fuel. It is considered
to be one of the more acutely toxic fractions of oil, with LC50s
for its water-soluble fraction ranging from 0.5 to 7 mg/L for
marine invertebrates [5,21]. The PAHs are considered to be
the primary source of persistent toxicity in number 2 oil. The
toxicological members of this fuel are naphthalenes, uorenes,
phenanthrenes, dibenzothiophenes, uoranthenes, and py-
renes, despite the fact that pyrene is usually associated with
a pyrolitic origin [6]. A recent survey on the presence of the
Prestige fuel oil in seawater indicated that naphthalene, phen-
anthrene, and their alkyl derivatives were the most abundant
hydrocarbons dissolved in water [22]. Thus, toxicity will be
addressed considering the above-mentioned parental PAHs and
alkyl derivatives for naphthalene and phenanthrene. The cy-
clopoid copepod O. davisae was chosen because it is repre-
sentative of probably the most abundant planktonic copepod
genus in the sea, commonly occurring in estuarine, coastal,
and open waters throughout most oceans [23] and, as such,
being found in areas characterized by frequent hydrocarbon
inputs from reneries, tanker operations, and accidental oil
spills. Thus, Oithona spp. are an important food source for
sh and macroinvertebrates living in coastal areas, and they
are sensitive to a variety of environmental contaminants [24].
Toxicity was assessed using acute responses, because death
or narcosis are important adverse effects [16] and are easily
comparable between species [811,13,25].
MATERIALS AND METHODS
Chemicals
Parental PAHs, including naphthalene, phenanthrene, uo-
rene, dibenzothiophene, and pyrene, were purchased from Sig-
ma-Aldrich (purity, 96%; St. Louis, MO, USA), whereas
alkyl PAH derivatives, including 1-methylnaphthalene, 1,2-
dimethylnaphthalene, 1-methylphenanthrene, and 3,6-dimeth-
ylphenanthrene, were obtained from ULTRA Scientic (purity,
97%; North Kingstown, RI, USA). Dosage, extraction, and
analysis of PAHs were conducted with high-performance liq-
uid chromatography (HPLC)grade solvents from Merck
(Darmstadt, Germany). All other chemicals were of analytical
grade and obtained from Sigma-Aldrich.
Experimental organisms and stock culture conditions
Adults of O. davisae were obtained from stock cultures
reared in 20-L metacrilate containers. The copepod cultures
are routinely kept with the heterotrophic dinoagellate Ox-
yrrhis marina as food, added at concentrations of 1,000 cells/
ml (0.1 mg C/L) [26]. Oxyrrhis marina, on its own, is grown
on the algae Rhodomonas salina. Cultures were maintained
at 20C under a 16:8-h light:dark photoperiod with a uores-
cent light (2040 E/m
2
).
Experimental design
Acute exposures to single or an equitoxic mixture combi-
nation of the studied PAHs were conducted as follows: Ex-
perimental treatments consisted of three replicates of a control
(no PAHs added) and four to seven test concentrations of the
studied PAHs. Acetone (HPLC grade, 0.5 ml/L) was used
as a carrier in all treatments. A control without acetone also
was included to ensure that baseline mortality was not affected
by acetone. Groups of 20 to 40 adult Oithona davisae were
incubated in 150-ml Pyrex glass bottles (Afora, Barcelona,
Spain) lled with 100 ml of the appropriate test solution. The
bottles were sealed with a Teon screw cap (Afora) and kept
standing. The copepods were exposed to PAHs without food
for 48 h. At the end of the exposure period, the contents of
the bottle were sieved through a submerged, 60-m nylon
mesh, and the copepods, which were retained by the mesh,
washed in ltered seawater and transferred to a Petri dish to
be checked for swimming activity under a stereomicroscope.
After a period of 1.5 to 3 h in clean seawater, animals were
examined again to assess the degree of recovery from nar-
cotization [19]. Animals that did not recover (i.e., were not
actively swimming) were considered to be dead. With this
procedure, we obtained an estimate of narcotic effects (animals
not swimming but alive at the end of the incubation) and lethal
effects (animals that did not recover mobility after the postin-
cubation period in ltered seawater without PAHs).
The combined toxicity of all but 3,6-dimethylphenathrene,
which was not toxic to O. davisae at concentrations less than
water-solubility levels [27], was determined using a xed-ratio
design. In such a design, the total concentration of the mixture
is varied while keeping the mixture ratio constant, so a com-
plete concentration (dose)response relationship of the mixture
can be described experimentally. A mixture with components
that were combined at the ratio of the median effect concen-
trations (EC50s) of the individual components was chosen.
This often-used design leads to the so-called equitoxic mix-
tures and is widely used for assessing the combined effect of
chemical mixtures [28].
Chemical analysis
To check whether the intended concentrations of the test
chemicals were achieved in the applied test solutions, dupli-
cated water-sample test solutions were collected from the ex-
perimental vessels at the beginning (t 0 h) and end (t 48
h) of test. Because of the large number of test concentrations
used, chemical analyses were restricted to one and three ex-
posure levels for single-substance and mixture-toxicity assays,
respectively. Exposure levels were selected to include nominal
concentrations close to the estimated EC50s for single-sub-
stance acute tests, whereas mixture-assay exposure levels in-
cluded 0.5, 1, and 1.5 toxic units (TUs). Polycyclic aromatic
hydrocarbons of test solutions were extracted from water sam-
ples and determined by HPLC using the method described by
Fernandez and Bayona [29] and by Barata et al. [16]. Test
solutions were extracted from water and preconcentrated in
solid-phase Bond Elut C18 extraction cartridges (Varian, Palo
2994 Environ. Toxicol. Chem. 24, 2005 C. Barata et al.
Alto, CA, USA). Hydrocarbon analytes were eluted from the
solid-phase cartridges with 5 ml of propan-2-ol:glacial acetic
acid:toluene:petroleum ether (0.9:0.1:2:7) at 40 to 60C. The
obtained solutions were then evaporated by gentle ushing
with N
2
nearly to dryness, and the residue was resuspended
in 1 ml of mobile-phase nanopure water:acetonitrile (1:1). The
resulting extracts were analyzed using a dual-pump HPLC
equipped with a UV diode-array detector fromShimadzu (Kyo-
to, Japan) using 220 nm for naphthalene, methyl- and dimeth-
ylnaphthalenes, and dibenzothiophene and 254 nm for phen-
anthrene, methyl- and dimethyl phenanthrene, anthracene,
uorene, uoranthene, and pyrene. The nal sample extract
(30 l) was injected onto a reversed-phase C
18
column (particle
size, 5 m; length, 150 mm; inner diameter, 4.6 mm; Vydac-
Grace, Philadelphia, PA, USA). Injection was carried out using
a Rheodyne valve (Rohnert Park, CA, USA) with a sample-
loop volume of 50 l tted to a Shimadzu autosampler. The
mobile phase used in the chromatographic separation consisted
of a binary mixture of acetonitrile and water at a constant ow
rate of 0.8 ml/min isocratically with 40% acetonitrile to sep-
arate naphthalene from a coelution. For all the other com-
pounds, a ow rate of 1 ml/min was used. Initial conditions
were 50% acetonitrile for 10 min, then linearly programmed
up to 65% acetonitrile for 30 min. Recoveries of individual
PAHs from water samples were determined from duplicated
standard solutions of known concentration, which were ex-
tracted and eluted at the same time as test solutions. Recoveries
(mean standard deviation [SD]) were analyte dependent,
being 52% 5%, 66% 10%, 75% 13%, and 93% 11%
for naphthalene, methylnaphthalene, dimethylnaphthalene, and
the rest of the studied PAHs, respectively. Quantication of
PAH analytes was conducted with eight-point calibration
curves relative to anthracene, which was used as an internal
surrogate. The correlation coefcients of the calibration curves
were always higher than 0.998.
Data analysis
Concentrationresponse curves were obtained for each
compound and for the mixture combination to derive EC50s
considering narcosis and lethality effects. The EC50s were
determined by tting observed responses to the nonlinear al-
losteric decay regression model according to method described
by Barata et al. [30]. Model accuracy was assessed by using
the adjusted coefcient of determination (r
2
) and by analyzing
residual distribution [31]. Signicance of the entire regression
and of regression coefcients was determined by analysis of
variance and t tests, respectively [31]. Curve-tting was per-
formed with the SPSS 11.2 statistical package (Chicago, IL,
USA) using the least-square method. Wherever the inspection
of residuals evidenced nonhomogeneity of the variance, the
estimates were corrected using the weighted least-square meth-
od, in which the weight was proportional to the reciprocal of
the variance.
To assess whether toxicity responses of O. davisae to single
PAHs were caused by nonpolar narcosis and, hence, could be
predicted from baseline QSAR models, estimated EC50s were
used to derive a QSAR model based on the log K
ow
values of
the studied hydrocarbons:
1
Log m log K a
owi
EC50
i
where EC50
i
(mol/L) is the median effect concentration for
compound i, a and m are the respective intercept and slope,
and K
owi
is the octanolwater partition coefcient for com-
pound i, which was obtained from existing data sets [8,11,27].
Using a large data set of acute toxicity responses (48-h LC50s)
of nonpolar narcotic chemicals to Daphnia magna, Hermens
et al. [14] and the Organisation for Economic Cooperation and
Development [32] reported regression coefcients of 0.95 and
4.85 for m and a, respectively. Furthermore, despite using
a slightly different equation, Swartz et al. [8] and Lee et al.
[11], studying the toxicity of PAHs to aquatic amphipods, also
found that lethal concentration levels followed log K
ow
-de-
pendent QSAR models with absolute slopes ranging from 0.97
to 1.33. Therefore, the toxicity of PAHs acting as nonpolar
narcotic chemicals in O. davisae should be expected to t the
log K
ow
-dependent QSAR model of Equation 1 with a slope
of approximately one.
Additivity of the equitoxic mixture of the studied PAHs
was tested by using the concentration addition concept. At a
mixture concentration that equals the sum of 1/n of the EC50s
of the individual components, concentration addition predicts
exactly 50% effect of the mixture:
n
c
i
1
ECx i1 i
where n is the number of mixture components, ECx
i
is the
effective concentration of the i-th mixture component that pro-
vokes x% effect when applied singly, and c
i
is the concentra-
tion of the respective component in the mixture. Each fraction
(c
i
/ECx
i
) represents the concentration of a mixture component
scaled for its relative toxicity and is generally termed the TU
of that component, which equals one at the 50% effect of the
mixture. Thus, equitoxic mixtures of PAHs with components
that act additively and are mixed at the ratio of their EC50s
should have an EC50 of 1 TU.
RESULTS
Chemical analyses
For single-substance toxicity tests, actual concentrations of
the studied PAHs in freshly prepared solutions were within 20
to 25% of the nominal aqueous solutions and decreased by 30
to 55% after 48 h (Table 1), with more volatile hydrocarbons
(naphthalene and methylnaphthalene) showing greater overall
decreasing levels than heavier and, hence, less volatile ones
(phenanthrene, dibenzothiophene, uoranthene, and pyrene).
When mixed together, individual PAH components at 1 TU
(Table 1) and at 0.5 and 1.5 TU (data not shown) revealed
equivalent decreasing patterns relative to those of single-sub-
stance exposures. In relation to this, actual PAH concentrations
in the remaining test solutions for single- and mixture-toxicity
assays were assumed to vary similarly when compared with
the measured solutions and were determined as the interpolate
midpoint value between the initial and nal concentrations.
Acute responses
In all bioassays, survival of O. davisae adults in control
treatments (mean SD, 94.8% 3.9%) did not differ sig-
nicantly (Students t test, p 0.05) from that of solvent
control treatments (93.3 3.9%). For all the tested PAHs,
acute responses (immobility because of narcosis or death) fol-
lowed the allosteric decay model, in which no or low response
at low concentration occurs, followed by an accelerating neg-
ative response as concentrations increase. In all cases, the re-
siduals of the regression model obtained were normally dis-
Single and mixture toxicity of PAHs to Oithona davisae Environ. Toxicol. Chem. 24, 2005 2995
Table 1. Nominal versus measured concentrations (mol/L) of the studied polycyclic aromatic hydrocarbons (PAHs) in test solutions freshly
prepared (t
0
) and at 48 h (t
48
)
a
Nominal
Measured concentrations
Mean t
0
SD t
0
Mean t
48
SD t
48
Single solutions
Naphthalene
1-Methylnaphthalene
1,2-Dimethylnaphthalene
Phenanthrene
1-Methylphenanthrene
54.61
31.62
6.39
3.37
3.12
41.61
25.12
5.47
3.8
2.41
4.08
1.51
0.52
0.05
0.01
28.57
15.20
3.54
2.31
1.75
6.10
2.13
0.26
0.06
0.04
3,6-Dimethylphenanthrene
Fluorene
Dibenzothiophene
Fluoranthene
Pyrene
0.48
12.03
3.8
0.99
0.74
0.48
10.3
3.66
0.96
0.64
0.04
1.70
0.32
0.00
0.02
0.22
7.39
1.89
0.47
0.5
0.02
0.6
0.03
0.02
0.05
PAH mixture
Naphthalene
1-Methylnaphthalene
1,2-Dimethylnaphthalene
Phenanthrene
1-Methylphenanthrene
6.24
2.07
0.44
0.33
0.54
4.28
1.23
0.42
0.24
0.45
0.51
0.06
0.07
0.01
0.02
2.93
1.11
0.31
0.2
0.39
0.05
0.00
0.04
0.01
0.01
Fluorene
Dibenzothiophene
Fluoranthene
Pyrene
1.20
0.33
0.11
0.08
0.91
0.33
0.09
0.08
0.00
0.00
0.02
0.00
0.82
0.27
0.07
0.06
0.01
0.01
0.01
0.00
a
Mean and standard deviation (SD) values for duplicated samples taken at the beginning and end of tests are reported. For clarity, results for
PAH mixtures are only reported for the equitoxic concentration of one toxic unit. The 3,6-dimethylphenanthrene concentration levels are not
reported for PAH mixture because it was not used as such.
tributed (KolmogorovSmirnov test, p 0.05), and the co-
efcients of determination, except that for methylphenanthrene
(r
2
0.64), were greater than 0.7 (Table 2). Acute toxicity
responses, measured as the lack of mobility after 48-h aqueous
exposures, varied dramatically across the studied PAHs, with
pyrene and naphthalene being the most and least toxic, re-
spectively, to O. davisae adults (Fig. 1). The PAHs were char-
acterized by exerting different degrees of narcotization to O.
davisae adults, as evidenced by steeper concentrationre-
sponse curves (Fig. 1, dotted lines) and lower EC50s relative
to those determined for survival responses (Table 2). Naph-
thalene, uoranthene, pyrene, and methylphenanthrene showed
the highest degree of narcotization, with narcosis accounting
for 29 to 37% of the observed acute effects. Alternatively,
phenanthrene, dimethylnaphthalene, dithiobenzene, and uo-
rene did not cause any narcotic effect on O. davisae adults.
QSARs for PAHs
Except for methylphenanthrene, acute responses of O. dav-
isae varied inversely with the K
ow
of the studied PAHs (Fig.
2). The obtained relationships between the EC50 and log K
ow
values, depicted in Table 3 both with and without considering
the outlier methylphenanthrene, were all signicant and ex-
plained more than 85% of the variability. Furthermore, the
estimated regression coefcients m and a (Table 3) did not
differ signicantly (Students t test, p 0.05) from those re-
ported (m 1, a 4.81) for baseline toxicity data in
Daphnia magna exposed to a large set of nonpolar narcotic
chemicals [14,32], thus suggesting common mechanisms of
PAH toxicity for both species.
Mixture toxicity
The EC50s for survival responses were used to establish
equitoxic mixtures of the studied PAHs, in which each indi-
vidual component contributed equally to the total toxicity. Di-
methylphenanthrene was excluded from the mixture, because
it did not cause toxic effect to O. davisae. Seven nominal
equitoxic concentrations (0, 0.25, 0.5, 1, 1.5, 2, and 2.5 TU,
which were further recalculated to account for actual concen-
trations of the individual components) were used to test the
additivity of PAH mixtures for narcotic and lethal effects (see
the second equation in Materials and Methods). Mixture tox-
icity curves also followed the allosteric decay model (Fig. 3
and Table 2), showing little narcotic effect on O. davisae
adults. Estimated EC50s considering narcotic and mortality
effects did not differ signicantly from 1 TU (Students t test,
p 0.05), thus denoting additivity of effects of PAH mixtures.
DISCUSSION
The present study evaluated acute responses of single and
mixture combinations of PAHs associated with the Prestige
oil spill to the copepod O. davisae. The results obtained in-
dicated that all but dimethylphenanthrene were toxic to O.
davisae at lower concentrations than their water-saturation lev-
els [10,27]. In ve of the nine studied PAHs that were toxic
to O. davisae, adults became narcotized, recovering shortly
after being transferred to clean seawater, with naphthalene and
uoranthene being the compounds exerting the greater effects.
Narcotization of water-soluble fuel oil extracts to zooplankton
species, particularly to the copepod E. afnis, has been as-
sociated previously with volatile components of the hydro-
carbon mixture (i.e., benzenes, toluenes, and xylenes) [19,33].
The results presented here show that more persistent com-
pounds, such as PAHs, can also narcotize copepods after 48-
h exposure at concentrations as much as 30% lower than those
impairing survival. Although not considered in the present
study, it is likely that copepod inactivity caused by partial or
complete narcotization severely impairs behavioral responses,
such as feeding and, hence, egg production, thus causing sub-
lethal effects on reproduction [16]. Furthermore, if narcosis
2996 Environ. Toxicol. Chem. 24, 2005 C. Barata et al.
Table 2. Log octanolwater partition coefcients (log K
ow
) values and toxic effects of single substance and polycyclic aromatic hydrocarbon
(PAH) mixtures to Oithona davisae adults
a
PAH Log K
ow
EC50 narcotization n r
2
EC50 survival n r
2
Naphthalene
1-Methylnaphthalene
1,2-Dimethylnaphthalene
Phenanthrene
1-Methylphenanthrene
3.37
3.87
4.31
4.57
5.14
34.93 (33.3636.51)
14.33 (13.2815.37)
4.06 (3.834.28)
3.58 (2.904.26)
3.42 (2.064.77)
18
18
18
27
27
0.97*
0.97*
0.99*
0.82*
0.76*
56.13 (42.7369.52)
18.65 (16.6420.65)
3.95 (3.664.24)
2.93 (2.473.39)
4.84 (2.257.43)
18
18
18
15
27
0.75*
0.99*
0.98*
0.95*
0.64*
3,6-Dimethylphenanthrene
Fluorene
Dibenzothiophene
Fluoranthene
Pyrene
Mixture
5.51
4.18
4.49
5.22
5.18

0.5
69.98 (7.9811.99)
2.81 (2.493.13)
0.66 (0.530.78)
0.53 (0.380.69)
1.02 (0.911.13)
32
30
18
36
36
21

0.75*
0.87*
0.84*
0.73*
0.97*
0.5
10.83 (8.7312.93)
2.99 (2.703.29)
0.97 (0.731.21)
60.76 (0.471.06)
1.09 (0.931.26)
32
15
18
15
15
21

0.88*
0.88*
0.84*
0.72*
0.95*
a
Estimated median effective concentration effect values (48-h EC50) for narcosis and survival responses were determined assuming an allosteric
decay nonlinear model. The 95% condence intervals for the EC50 are shown in parentheses. The EC50 values are reported as mol/L for
single-substance toxicity tests and in toxic units for the PAH mixture. Log K
ow
values are from Swartz et al. [8], Lee et al. [11], and Mackay
[27]. The 3,6-dimethylphenathrene was not toxic at concentrations exceeding its estimated water solubility. r
2
coefcient of determination;
n sample size; missing value. *p 0.05.
Table 3. Quantitative structureactivity relationship models obtained
for the studied polycyclic aromatic hydrocarbons in Oithona davisae
adults considering the data shown in Table 2 and Equation 1
a
Effect a m n r
2
Considering the nine PAHs
Narcotization
Survival
4.43
4.49
0.85
0.85
9
9
0.85*
0.82*
Excluding methylphenanthrene
Narcotization
Survival
4.49
5.10
0.97
1.01
8
8
0.97*
0.96*
a
a regression intercept; m regression slope; n sample size; r
2
coefcient of determination.
Fig. 1. Survival and narcotizing responses of Oithona davisae to
single exposures of naphthalene (N), methylnaphthalene (N1), di-
methylnaphthalene (N2), phenanthrene (P), methylphenanthrene
(P1), uorene (Fl), dibenzothiophene (DBT), uoranthene (Ft), and
pyrene (Py). Data have been tted to an allosteric decay model.

survival; narcotization.
persists because of longer exposure periods, the copepod will
eventually die.
Except for alkylated phenanthrenes, toxicity of single PAHs
was inversely related (r
2
95%) to their K
ow
values, following
the log K
ow
-dependent QSAR models derived for baseline tox-
icity of nonpolar narcotic chemicals to D. magna [14,32].
Indeed, the regression QSAR models obtained considering ei-
ther narcotizing or survival responses explained more than
95% of the observed variability, and the regression coefcients
obtained (m 0.97 and 1.01, respectively, and a 4.49
and 5.10, respectively, where m and a are the slope and
intercept) matched those estimated for D. magna toxicity data
of nonpolar narcotic chemicals (m 1, a 4.81). Swartz
et al. [8] and Lee et al. [11], studying toxicity of PAHs to
aquatic amphipods, also found that lethal concentration levels
followed log K
ow
-dependent QSAR models with absolute
slopes ranging from 0.97 to 1.33. This indicates that in the
absence of photoactivation, toxicity of low-molecular-weight
PAHs to O. davisae and amphipod species mainly results from
nonpolar narcosis and, hence, can be predicted with condence
by K
ow
-dependent QSAR models derived for predicting base-
line toxicity [14,32]. It is important to consider that K
ow
QSAR
models are empirical regression equations; hence, for similar-
acting chemicals, their elevations are expected to vary across
species, exposure scenarios, and end points [11,14]. In the
present study, despite using a different species, the use of end
points (48-h LC50 of water-only exposures) equivalent to those
used in D. magna allowed similar regression parameters to be
obtained, thus suggesting that PAH bioavailability and toxicity
to freshwater cladocerans and marine copepods may be similar.
The QSAR models derived for amphipod species, however,
denoted greater sensitivity to PAHs [8,11,25]. Nevertheless,
the use of different exposure scenarios (sediment) and/or lon-
ger exposure periods (1014 d) precluded a reliable compar-
ison with the present results. According to Lee et al. [11],
uptake and toxicity of PAHs increase with exposure time; thus,
the use of longer exposure periods should have resulted in
greater sensitivity.
Despite numerous studies that have evaluated the toxicity
Single and mixture toxicity of PAHs to Oithona davisae Environ. Toxicol. Chem. 24, 2005 2997
Fig. 2. Relationships between acute responses of Oithona davisae
depicted as median effect concentrations (EC50s) and the log octanol
water partition coefcients (log K
ow
) of the studied polycyclic aromatic
hydrocarbons. Estimated quantitative structureactivity regression
curves considering all but methylphenanthrene results also are de-
picted. Error bars represent 95% condence intervals. The EC50 are
depicted in a log inverse scale. See Figure 1 for abbreviations.

survival; narcotization.
Fig. 3. Survival and narcotizing responses of Oithona davisae to an
equitoxic mixture of nine polycyclic aromatic hydrocarbons, including
naphthalene, methylnaphthalene, dimethylnaphthalene, phenanthrene,
methylphenanthrene, uorene, dibenzothiophene, uoranthene, and
pyrene. Data have been tted to an allosteric decay model.

survival; narcotization.
of water-soluble oil or hydrocarbon extracts on different co-
pepod species [1], the literature contains little information re-
garding the toxicity of individual hydrocarbons to copepods
with which the present results involving O. davisae may be
compared. Ott et al. [18], studying acute toxicity of naphtha-
lene and three alkylated derivatives to the estuarine copepod
E. afnis, found a logarithmic decrease in toxic concentration
with increasing methylation, with reported sensitivities (24-h
LC50) for naphthalene, methylnaphthalene, and dimethylnaph-
thalenes of 30, 10, and 5 mol/L, respectively. Barata et al.
[16], studying lethal and sublethal responses of the meio-
benthic copepod Tisbe battagliai exposed to aqueous uor-
anthene levels, reported a 48-h LC50 for adult stages of 0.5
mol/L. Comparison of the previously mentioned values with
those obtained in the present work denoted equivalent levels
of sensitivity across copepod species. For other aquatic crus-
tacea, including amphipod and estuarine shrimp species, re-
ported acute sensitivities to aqueous concentrations of single
PAHs were, in most cases, similar to those obtained for O.
davisae, with LC50s ranking from 10 to 27.3, 6.3 to 12.1, 0.5
to 4.5, 0.5 to 1.3, 1.6 to 3.2, 0.1 to 0.3, and 0.07 to 0.3 mol/
L for naphthalene, methylnaphthalene, dimethylnaphthalene,
phenanthrene, uorene, uoranthene, and pyrene, respectively
[11,21,25]. Of special interest are the unexpectedly decreased
levels of toxicity found with increasing methylation for the
alkylated derivatives of phenanthrene. Contrary to the present
results, recent studies have found high toxicity levels of alkyl
phenanthrenes to sh embryos [34,35]. Acute toxicity of alkyl
phenanthrene to sh embryos has been related to oxidative
stress and effects on cardiovascular morphogenesis [34]. For
other studied organisms, however, nonpolar narcosis is con-
sidered to be the common mechanism of toxicity of PAHs and
alkylated derivatives [8,9,11,14,25]. Therefore, differences in
the bioavailability and elimination of alkylated relative to pa-
rental PAHs are likely to be involved in their differential tox-
icity to most organisms. Bioaccumulation studies of sh and
invertebrate species [21,36,37] have indicated that alkylated
naphthalenes are characterized by accumulating both more rap-
idly and to a greater extent than naphthalene and that they are
released more slowly. On the contrary, lower bioconcentration
levels or decreased bioavailability followed by an efcient
metabolism of alkylated phenanthrenes relative to phenan-
threne have been reported in sh and mussels. Therefore, great-
er and lower concentration levels at equilibrium within co-
pepods may explain the greater and lower toxicity of alkylated
derivatives of naphthalenes and phenanthrenes, respectively
[36,37]. Unfortunately, to our knowledge, no studies have
compared bioaccumulation and toxicity patterns of alkylated
and parental PAHs in marine copepod species.
Petrogenic PAHs are characterized to occur in complex
mixtures [2,6]; thus, environmental risk assessment procedures
derived for PAHs should focus on mixture toxicity. For PAHs,
which are distinctively narcotic compounds, mixture toxicity
effects typically are considered to be additive [8,9,14,25]. Nev-
ertheless, several studies have reported exceptions to the ad-
ditivity effects model. Synergic effects were found in amphi-
pods for mixtures including photoactivated PAH compounds
[10] and in sh embryos for alkyl phenanthrenes [34,35]. Al-
ternatively, less-than-additive effects have been reported for
sediment PAH mixtures, including high-molecular-weight and
high-K
ow
compounds that are not toxic in single exposures [11].
The above-mentioned results suggest that mixture toxicity de-
pends on the environmental conditions and the nature of the
PAH involved. Marine zooplanktonic organisms are likely to
be affected only by the most soluble low-molecular-weight
PAH compounds, because heavy congeners tend to be parti-
tioned into sediment or seston particles or into biological tis-
sues [2,22]. In such conditions, the present results indicate that
mixture toxicity of PAHs showed a clear, concentration ad-
ditive effect, with an estimated mean (95% CL) equitoxic mix-
ture toxicity of 1.09 (0.931.26) TU. It is interesting to note
that when considering narcotization effects, the estimated equi-
toxic mixture effects were almost identical to the values re-
ported above, thus indicating that in complex mixtures of com-
pounds with and without narcotization, the effects level out
to produce no narcotic effects.
Finally, it should be interesting to compare acute concen-
2998 Environ. Toxicol. Chem. 24, 2005 C. Barata et al.
tration effects obtained for O. davisae (48-h LC50) exposed
to individual hydrocarbons with levels that might be encoun-
tered in the marine environment. Background levels less than
1 and 0.043 g/L for dissolved hydrocarbons and naphtha-
lenes, respectively, have been reported in several studies
[38,39], whereas dissolved hydrocarbon levels as high as 5.4
mg/L and naphthalene levels of 0.86 mg/L were found 3 d
after a slick of fresh number 2 fuel oil [40]. However, in most
cases, PAH levels never exceed 60 g/L following an oil spill
[1,39]. Nevertheless, in the vicinity of an oil spill, complex
mixtures of hundreds of different hydrocarbons are common
[2,6]; thus, even at low exposure levels, single PAHs may act
additively, impairing survival of copepod populations. In the
present study, the data depicted in Table 2 (converted to mg/
L) indicated that observed narcotic and lethal concentration
level effects of O. davisae adults exposed to single naphthalene
solutions (EC50 4.467.19 mg/L) were in excess of those
in the marine environment, even in the immediate vicinity of
an oil spill. However, in a complex mixture of only nine com-
pounds, exposure levels of 0.58 mg/L of naphthalene and 1.16
mg/L of the sum of the nine studied hydrocarbons were able
to severely impair survival of O. davisae, thus indicating that
the acute effects obtained for PAH mixtures were somewhat
closer to those that might only occur near an oil spill [40].
AcknowledgementThe present study was funded by the Spanish
Ministerio de Educacion y Ciencia and the Spanish projects
VEM2003-20037 and REN 2003-06917-CO
2
-O2. C. Barata and A.
Calbet were supported by a Ramon y Cajal contract from the Min-
isterio de Educacion y Ciencia. E. Vergara provided technical assis-
tance.
REFERENCES
1. Corner EDS. 1978. Pollution studies with marine plankton. Part
I. Petroleum hydrocarbons and related compounds. Adv Mar Biol
15:259380.
2. Albaiges J. 1989. Marine Pollution. Editorial Hemisphere, New
York, NY, USA.
3. Pauly D, Christensen V, Guenette S, Pitcher J, Sumaila UR, Wal-
ters CJ, Watson R, Zeller D. 2002. Towards sustainability in world
sheries. Nature 418:689695.
4. Hoong Gin KY, Huda MK, Lim WK, Tkalich P. 2001. An oil
spillfood chain interaction model for coastal waters. Mar Pollut
Bull 42:590597.
5. Neff JM, Anderson JW, Cox BA, Laughlin RB, Rossi SS, Tatem
HE. 1976. Effects of petroleum on survival, respiration, and
growth of marine animals. Proceedings, Sources, Effects and
Sinks of Hydrocarbons in the Aquatic Environment, Washington,
DC, August 911, pp 515539.
6. Pastor D, Sanchez J, Porte C, Albaiges J. 2001. The Aegean Sea
oil spill in the Galicia Coast (NW Spain). I. Distribution and fate
of crude oil and combustion products in subtidal sediments. Mar
Pollut Bull 42:895904.
7. Van Brummelen TC, van Hattum B, Crommentuijn T, Kalf DF.
1998. Bioavailability and ecotoxicity of PAHs. In Neilson AH,
ed, The Handbook of Environmental Chemistry. Springer-Verlag,
Berlin, Germany, pp 203263.
8. Swartz RC, Schults DW, Ozretich JO, Lamberson JO, Cole FA,
DeWitt TH, Redmond MS, Ferraro SP. 1995. PAH: A model to
predict the toxicity of eld-collected marine sediment contami-
nated by polynuclear aromatic hydrocarbons. Environ Toxicol
Chem 14:19771987.
9. Swartz RC, Ferrando SP, Lamberson JO, Cole FA, Ozretich RJ,
Boese BL, Schults DW, Behrenfeld M, Ankley GT. 1997. Pho-
toactivation and toxicity of mixtures of polycyclic aromatic hy-
drocarbon compounds in marine sediment. Environ Toxicol Chem
16:21512157.
10. Boese BL, Ozretich RJ, Lamberson JO, Swartz RC, Cole FA,
Pelletier J, Jones J. 1999. Toxicity and phototoxicity of mixtures
of highly lipophilic PAH compounds in marine sediment: Can
the model be extrapolated? Arch Environ Contam Toxicol 36:
270280.
11. Lee JH, Landrum PF, Field LJ, Koh CH. 2001. Application of a
polycyclic hydrocarbon model and a logistic regression model
to sediment toxicity based on a speciesspecic, water-only
LC50 toxic unit for Hyalella azteca. Environ Toxicol Chem 20:
21022113.
12. Landrum PF, Lotufo GR, Gossiaux DC, Gedeon ML, Lee JH.
2003. Bioaccumulation and critical body residue of PAHs in the
amphipod, Diporeia spp.: Additional evidence to support toxicity
additivity for PAH mixtures. Chemosphere 51:481489.
13. Djomo JE, Dauta A, Ferrier V, Narbone JF, Monkiedje A, Njine
T, Garrigues P. 2004. Toxic effects of some major polyaromatic
hydrocarbons found in crude oil and aquatic sediments on Sce-
nedesmus subspicatus. Water Res 38:18171821.
14. Hermens JLM, Canton H, Janssen P, de Jong R. 1984. Quantitative
structureactivity relationships and toxicity studies of mixtures
of chemicals with anesthetic potency: Acute and sublethal toxicity
to Daphnia magna. Aquat Toxicol 5:143154.
15. Zaret TM, Suffern JS. 1976. Vertical migration in zooplankton
as a predator avoidance mechanism. Limnl Oceanogr 21:804
813.
16. Barata C, Baird DJ, Medina M, Albalat A, Soares AMVM. 2002.
Determining the ecotoxicological mode of action of toxic chem-
icals in meiobenthic marine organisms: Stage-specic short tests
with Tisbe battagliai. Mar Ecol Prog Ser 230:183194.
17. Lotufo GR. 1997. Toxicity of sediment-associated PAHs to an
estuarine copepod: Effects on survival, feeding, reproduction and
behavior. Mar Environ Res 44:149166.
18. Ott FS, Harris RP, OHara SC. 1978. Acute and sublethal toxicity
of naphthalene and three methylated derivatives to the estuarine
copepod, Euritemora afnis. Mar Environ Res 1:4957.
19. Berdugo V, Harris RP, OHara SC. 1977. The effect of petroleum
hydrocarbons on reproduction of an estuarine planktonic copepod
in laboratory cultures. Mar Pollut Bull 8:138143.
20. Fernandez-Latorre SR. 2003. La Huella del Fuel. Ensayos Sobre
el Prestige. Fundacion Santiago Rey Fernandez-Latorre, A Co-
runa, Spain.
21. Anderson JW, Neff JM, Cox BA, Tatem HE, Hightower GM.
1974. The effects of oil on estuarine animals: Toxicity, uptake
and depuration, respiration. In Vernberg FJ, Vernberg WB, eds,
Pollution and Physiology of Marine Organisms. Academic, New
York, NY, USA, pp 285310.
22. Albaiges J, Bayona JM. 2003. El Fuel. In Fernandez-Latorre SR,
ed, La Huella del Fuel. Ensayos Sobre el Prestige. Fundacion
Santiado Rey Fernandez-Latorre, A Coruna, Spain, pp 80103.
23. Galliene CP, Robins DB. 2001. Is Oithona the most important
copepod in the worlds oceans? J Plankton Res 23:14211432.
24. Willis KJ, Ling N. 2004. Toxicity of the aquaculture pesticide
cypermethrin to planktonic marine copepods. Aquat Res 35:263
270.
25. Landrum PF, Lotufo GR, Gossiaux DC, Gedeon ML, Lee JH.
2003. Bioaccumulation and critical body residue of PAHs in the
amphipod,Diporeia spp.: Additional evidence to support toxicity
additivity for PAH mixtures. Chemosphere 51:481489.
26. Saiz E, Calbet A, Broglio E. 2003. Effects of small turbulence
on copepods: The case of Oithona davisae. Limnol Oceanogr 18:
13041311.
27. Mackay D. 1982. Correlation of bioconcentration factors. Environ
Sci Technol 16:274278.
28. Konemann H. 1980. Structureactivity relationship and additivity
in sh toxicities of environmental pollutants. Ecotoxicol Environ
Saf 4:415421.
29. Fernandez P, Bayona JM. 1992. Use of off-line gel permeation
chromatography normal-phase liquid chromatography for the de-
termination of polycyclic aromatic compounds in environmental
samples and standard reference materials (air particulate matter
and marine sediment). J Chromatogr 625:141149.
30. Barata C, Baird DJ, Medina M, Albalat A, Soares AMVM. 2002.
Determining the ecotoxicological mode of action of toxic chem-
icals in meiobenthic marine organisms: Stage-specic short tests
with Tisbe battagliai. Mar Ecol Prog Ser 230:183194.
31. Zar JH. 1996. Bioestatistical Analysis. Prentice-Hall Internation-
al, Upper Saddle River, NJ, USA.
32. Organisation for Economic Cooperation and Development. 1995.
Guidance Document for Aquatic Effects Assessment. OECD En-
vironment Monographs 92. Paris, France.
Single and mixture toxicity of PAHs to Oithona davisae Environ. Toxicol. Chem. 24, 2005 2999
33. Lee RF. 1975. Fate of petroleum hydrocarbons in marine zoo-
plankton. Proceedings, Joint Conference on Prevention and Con-
trol of Oil Spills, San Francisco, CA, USA, March 2527, pp
549553.
34. Brinkworth LC, Hodson PV, Tabash S, Lee P. 2003. CYP1A in-
duction and blue sac disease in early developmental stages of
rainbow trout (Oncorhynchus mykiss) exposed to retene. J Tox-
icol Environ Health A 66:627646.
35. Barron MG, Carls MG, Heintz R, Rice SD. 2004. Evaluation of
sh early life-stage toxicity models of chronic embryonic ex-
posures to complex polycyclic aromatic hydrocarbon mixtures.
Toxicol Sci 78:6067.
36. Rantamaki P. 1997. Release and retention of selected polycyclic
aromatic hydrocarbons (PAH) and their methylated derivatives
by the common mussel (Mytilus edulis) in the brackish water of
the Baltic sea. Chemosphere 35:487502.
37. Jonsson G, Bechmann RK, Bamber SD, Baussant T. 2004. Bio-
concentration, biotransformation, and elimination of polycyclic
aromatic hydrocarbons in sheepshead minnows ( Cyprinodon var-
iegatus) exposed to contaminated seawater. Environ Toxicol
Chem 23:15381548.
38. Brown RA, Searl TD, Elliot JJ, Phillips BG, Brandon DE, Mon-
aghan PH. 1973. Distribution of heavy hydrocarbons in some
Atlantic Ocean waters. Proceedings, Joint Conference on Pre-
vention and Control of Oil Spills, American Petroleum Institute,
Washington, DC, March 1315, pp 505519.
39. Fatima M, Meniconi G, Ganardo IT. 2002. Brazilian oil spills
chemical characterizationCase studies. Environmental Foren-
sics 3:303321.
40. Frankenfeld JW. 1973. Factors governing the fate of oil at sea:
Variations in the amounts and types of dissolved or dispersed
materials during the weathering process. Proceedings, Joint Con-
ference on Prevention and Control of Oil Spills, Washington, DC,
March 1315, pp 485495.