Environmental Toxicology and Chemistry, Vol. 25, No. 9, pp. 2423–2431, 2006
? 2006 SETAC
Printed in the USA
0730-7268/06 $12.00 ? .00
CHRONIC TOXICITY OF POLYCYCLIC AROMATIC COMPOUNDS TO THE
SPRINGTAIL FOLSOMIA CANDIDA AND THE ENCHYTRAEID
STEVEN T.J. DROGE,†‡ MIRIAM LEO´N PAUMEN,†§ ERIC A.J. BLEEKER,†§ MICHIEL H.S. KRAAK,*§ and
CORNELIS A.M. VAN GESTEL†
†Department of Animal Ecology, Institute of Ecological Science, Vrije Universiteit, De Boelelaan 1085, 1081 HV Amsterdam,
‡Institute for Risk Assessment Sciences, Utrecht University, Yalelaan 2, 3584 CL Utrecht, The Netherlands
§Department of Aquatic Ecology and Ecotoxicology, Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam,
Kruislaan 320, 1098 SM Amsterdam, The Netherlands
(Received 8 November 2005; Accepted 20 February 2006)
Abstract—An urgent need exists for incorporating heterocyclic compounds and (bio)transformation products in ecotoxicological
test schemes and risk assessment of polycyclic aromatic compounds (PACs). The aim of the present study therefore was to determine
the chronic effects of (heterocyclic) PACs on two terrestrial invertebrates, the springtail Folsomia candida and the enchytraeid
Enchytraeus crypticus. The effects of 11 PACs were determined in chronic experiments using reproduction and survivalasendpoints.
The results demonstrated that as far as narcosis-induced mortality is concerned, effects of both homocyclic and heterocyclic PACs
are well described by the relationship between estimated pore-water 50% lethal concentrations and log Kow. In contrast, specific
effects on reproduction varied between species and between compounds as closely related as isomers, showing up as deviations
from the relationship between pore-water 50% effect concentrations and log Kow. These unpredictable specific effects on reproduction
force one to test the toxicity of these PACs to populations of soil invertebrates to obtain reliable effect concentrations for use in
risk assessment of PACs.
Keywords—Chronic toxicityPolycyclic aromatic compounds
Folsomia candidaEnchytraeus crypticus
Contamination with polycyclic aromatic compounds
(PACs) often consists of a variety of different compounds,
including hydroxylated, oxygenated, and chlorinated com-
pounds as well as nitro- and amino-PACs. In addition, many
heterocyclic compounds containing in-ring substitutions of ni-
trogen, sulfur, and/or oxygen atoms have been identified in
such emissions , yet, current risk assessment for PACs fo-
cuses on homocyclic compounds only. Hence, an urgent need
exists for incorporating heterocyclic PACs and (bio)trans-
formation products (oxidation products) in ecotoxicological
test schemes and risk assessment of PACs. Therefore, in ad-
dition to homocyclic compounds, the present study focused
on azaarene analogues and stable azaarene metabolites. Azaar-
enes (i.e., PACs in which one carbon atom has been replaced
by a nitrogen atom) are present in the environment in amounts
up to 1 to 10% of those of their homocyclic analogues .
Apart from their natural origin—for example, as alkaloids
—and the usual sources of PACs, basic azaarene structures
occur as moieties of pharmaceuticals [4,5] and pesticides [6,7].
Special attention will be paid to isomerism, because it has
been demonstrated that such slight differences in chemical
structure can result in substantial differences in toxicity [8–
Because soils and sediments are a major sink of PACs, the
biota in these compartments should be protected by well-de-
* To whom correspondence may be addressed
fined maximum permissible concentrations. Moreover, ade-
quate understanding of the environmental risks of PACs is
required, because bioremediation of polluted soils is expensive
and the most affected sites need to be prioritized. Relevant
toxicity data for soil organisms are scarce, though. In a recent
review concerning the toxicity of PACs to terrestrial inver-
tebrates, only 19 literature sources were mentioned . Com-
parisons between these few studies are complicated because
of the wide variety of test conditions and experimental setups.
The aim of the present study therefore was to determine the
long-term effects of PACs on two terrestrial invertebrates be-
longing to different phyla, the springtail Folsomia candida
and the enchytraeid (oligochaete) Enchytraeus crypticus.
These two species presumably are exposed in different ways
to PACs, because the springtail inhabits soil pores whereas the
oligochaete perturbates the soil. Springtails also are in close
contact with pore water through their ventral tube. Testing the
parthenogenetic springtail F. candida allowed comparison
with a closely related, sexually reproducing species, Folsomia
fimetaria, for which toxicity data have become available [15–
18]. The effects of the PACs were determined in chronic ex-
periments using reproduction and survival as endpoints.
MATERIALS AND METHODS
The parthenogenetic collembolan F. candida is a com-
monly used test species in soil toxicity tests . Recently, a
standardized test method to evaluate effects of toxicants on
enchytraeid survival and reproduction has been developed
Environ. Toxicol. Chem. 25, 2006S.T.J. Droge et al.
[20,21]. Both species are representatives of ecologically im-
portant functional groups and can reach high densities in var-
ious soil types [22,23]. They have relatively short life cycles,
can be cultured easily in the laboratory, and are suitable for
testing in both artificial and natural soils. Cultures of both
species have been maintained at Vrije Universiteit for several
years. The cultures were kept at 16?C under a 16:8-h light:
The selected PACs (Table 1) included six homocyclic com-
pounds: Naphthalene, the isomers phenanthrene and anthra-
cene, benz[a]anthracene, pyrene, and benzo[a]pyrene. The
chosen azaarenes quinoline and the isomers acridine and phen-
anthridine are analogues of the two- and three-ringed homo-
cycles. The two azaarene metabolites 9(10H)-acridone and
6(5H)-phenanthridinone also are isomers. Several properties
of these compounds are listed in Table 1. All compounds were
purchased from Sigma-Aldrich (Steinheim, Germany) except
for anthracene and benz[a]anthracene (Janssen Chimica,
The toxicity tests were performed using a standardized nat-
ural soil (Landwirtschaftliche Untersuchungs- und Forschung-
sanstalt [LUFA] 2.2, Speyer, Germany) sieved at 2 mm. This
soil, characterized as a sandy loam (particle size distribution:
50–2,000 ?m, 75.3%; 2–50 ?m, 16.6%; and ?2 ?m, 8.1%),
was taken from a meadow that had been free of pesticide use
or organic fertilization for more than four years. The total
organic carbon content was 2.3% ? 0.2%, and soil pH (0.01
M CaCl2) was 5.6 ? 0.4. Density, water-holding capacity, and
cation-exchange capacity were 1.15 kg/dm3, 46% (w/w), and
11 mval/100 g, respectively. The soil was kept at room tem-
perature at a moisture content of 5% until used in the tests.
The soil was spiked with the following nominal concen-
tration ranges of the selected compounds: For F. candida:
naphthalene, anthracene, benz[a]anthracene, benzo[a]pyrene,
and acridine at 62.5, 125, 250, 500, and 1,000 mg/kg; quinoline
and phenanthridine at 31, 62.5, 125, 250, 500, and 1,000 mg/
kg; pyrene at 15, 31, 62.5, 125, 250, 500, and 1,000 mg/kg;
phenanthrene at 12.5, 25, 50, 100, and 200 mg/kg; and acridone
and phenanthridinone at 10, 100, and 500 mg/kg; for E. cryp-
ticus: naphthalene, anthracene,
benz[a]anthracene, benzo[a]pyrene, quinoline, and acridine at
62.5, 125, 250, 500, and 1,000 mg/kg; phenanthridine at 31,
62.5, 125, 250, 500, and 1,000 mg/kg; and acridone and phen-
anthridinone at 10, 100, and 500 mg/kg. Controls and solvent
controls were included. Five replicates per concentration were
used for the homocyclic compounds and the azaarenes, and
four replicates per concentration were used for the transfor-
mation products. Acetone (purity, 99.8%; Riedel-de Hae ¨n,
Seelze, Germany) was used as a carrier solvent, and equal
volumes of acetone were added to all treatments. The PAC
solution was added to a quarter of the soil sample, and the
spiked soil was left overnight to allow the acetone to evaporate.
The next day, the rest of the soil was added, water was added
up to 46% ? 2% of the water-holding capacity, and the soil
was homogeneously mixed. Soil PAC concentrations and pH
were measured at the start of the experiment, after 10 d, and
at the end of the experiment (28 d). To determine actual PAC
concentrations and soil pH after 10 d, one additional replicate
was prepared for the highest and second-lowest test concen-
trations, and one additional replicate was prepared for all con-
centrations to determine actual PAC concentrations and soil
pH at the end of the test. The pH (KCl; average ? standard
deviation) was 5.7 ? 0.37 at the start of the experiment and
5.3 ? 0.34 at the end of the experiment (28 d).
Slight adaptations of the international guidelines for tox-
icity tests with F. candida [19,22] and E. crypticus  made
it possible to test these organisms under closely comparable
test conditions using the same series of PAC concentrations.
The tests started with either F. candida that were synchronized
to 11 to 13 d of age or adults of E. crypticus of approximately
0.4 to 0.6 cm that were adapted to clean LUFA 2.2 soil for 1
d. Each replicate consisted of 30 g (wet wt) spiked soil, con-
taining 10 individuals, in 100-ml glass jars. These jars were
closed off with black-plastic screw tops for the springtails and
with perforated aluminum foil for the enchytraeids. All glass-
ware used in the tests and in the PAC analysis, as well as the
screw tops, were cleaned with 99.8% acetone and 96% meth-
anol (both from Riedel-de Hae ¨n). At the start of the tests, 2
mg of food were added to each replicate on top of the soil
(dried yeast grains for F. candida and crushed oatmeal for E.
crypticus). To control fungal growth, extra food was added
only when necessary, and loss of water through evaporation
was compensated for by adding demineralized water once a
week. Test containers were kept at 20?C and 80% humidity
under a 16:8-h light:dark photoperiod.
The animals were extracted from the soil after the 28-d
exposure period. The soils with F. candida were gently stirred
with 100 ml of water, causing the surviving adults and the
produced juveniles to float on the surface of the suspension.
A digital photograph from this surface facilitated the counting
of individuals with use of a colony counter (Bel-Arts Products,
Pequannock, NJ, USA). Adult and produced juvenile enchy-
traeids were fixated by adding 5 ml of 96% ethanol to the soil.
After a few seconds, the soil was rinsed out of the jar and into
a plastic cup using 100 ml of tap water. Another 5 ml of 96%
ethanol was added, and the suspension was gently stirred. A
few drops of rose bengal dissolved in ethanol (1%) were added,
and the cup, closed off with a lid, was shaken rigorously for
5 s. After leaving this suspension overnight at 4?C, the bright
pink–colored enchytraeids were sieved over 160 ?m and
counted in a white, 80- ? 50-cm photo tray. As a result of
using adult, probably egg-bearing enchytraeid worms at the
start of the test, the first-born juvenile worms reached the adult
stage before the end of the 28-d exposure period. Therefore,
adults with a body length of 8 mm or less were considered to
be hatchlings of the original parents.
Actual PAC concentrations were determined in single soil
samples at the start of the experiment, after 10 d, and at the
end of the experiment (28 d). From these samples, 5 g of moist
soil were mixed with 5 g of anhydrous sodium sulfate (p.a.
Merck, Darmstadt, Germany) and Soxhlet-extracted in hexane
(purity, ?97%; Biosolve, Valkenswaard, The Netherlands) or
in acetonitrile (in case of the metabolites; analyzed high-per-
formance liquid chromatography reagent; purity, ?99.9%; J.T.
Baker, Deventer, The Netherlands) for 5 h using 33- ? 94-
mm cellulose extraction thimbles (Schleicher & Schuell, Das-
Chronic effects of PACs on soil invertebrates
Environ. Toxicol. Chem. 25, 2006 2425
Table 1. Selected test compounds and their Chemical Abstracts Service (CAS) number, molecular weight (MW), purity, log Kow, log Koc, and
water solubility (Sw)
Compound Structure CAS no.MW Purity (%)Log Kow
Homocyclic polycyclic aromatic compounds (PACs)
Phenanthrene 85-01-8178.2399.5 4.48a
Pyrene129-00-0 202.26 984.84a
Benz[a]anthracene 56-55-3 228.2999 5.54a
Benzo[a]pyrene 50-32-8 252.31 976.02a
Heterocyclic PACs (azaarenes)
Phenanthridine 229-87-8179.22 99?
9(10H)-acridone578-95-0 195.2299 2.95d
6(5H)-phenanthridinone 1015-89-0 195.2297 2.70f
aExperimental values from Helweg et al. .
bExperimental values from Jonassen et al. .
cPearlman et al. .
dExperimental values from Thomsen .
eCalculated from WSKOW (1.40; Syracuse Research Corporation, North Syracuse, NY, USA).
sel, Germany). Cleanup of the hexane samples using Baker-
bond spe? Columns (Silica gel; J.T. Baker, Phillipsburg, NJ,
USA) was performed in tests with anthracene and ben-
zo[a]pyrene (both with F. candida and E. crypticus) and phen-
anthrene and benz[a]anthracene (both with F. candida only)
but was found to be redundant for all other tests, because most
PAC concentrations were sufficiently high. Moreover, azaar-
enes were not released from the column packing. Either di-
rectly from the Soxhlet flasks or after the cleanup procedure,
1.5 ml of the hexane extracts was added to 2 ml of acetonitrile
(purity, ?99.8%; J.T. Baker, Deventer, The Netherlands). The
PAC was collected in the acetonitrile by blowing off thehexane
using a gentle stream of nitrogen. These samples were then
analyzed using a high-performance liquid chromatographic
system consisting of a Vydac 201TP reverse-phase column
(C18; length, 250 mm; inner diameter, 4.6 mm; film thickness,
5 ?m) with a Vydac 201GD guard column (R-P C18; length,
10 mm; inner diameter, 4.6 mm; film thickness, 10 ?m) con-
nected to a fluorescence detector (model FP-1520; Jasco, Es-
sex, UK) and a UVD320s Ultraviolet diode-array detector
Environ. Toxicol. Chem. 25, 2006S.T.J. Droge et al.
(Gynkotek, Germering, Germany). When peaks were below
detection levels, the hexane samples were concentrated up to
30-fold using Kuderna-Danish solvent evaporators. All vol-
umes were checked by weighing on a 0.1-mg microbalance
(R160P Sartorius, Go ¨ttingen, Germany), and samples were di-
luted with acetonitrile when necessary. In a separate experi-
ment, PAC-spiked LUFA 2.2 soil samples were used to cal-
culate extraction procedure recoveries and correction values
for all tested PACs in the total extraction process.
The concentrations at which 50 or 10% mortality was ob-
served (LC50 and LC10, respectively) and at which the re-
productive output was reduced by 50 or 10% compared to the
control (EC50 and EC10, respectively) were calculated ac-
cording to the method described by Haanstra et al. . Using
SPSS? (Ver 10.0.5; SPSS, Chicago, IL, USA), the following
logistic curve was fitted through the concentration–response
plot using actual initial concentrations:
y ? c/(1 ? e
where y is the effect parameter (survival or reproduction after
28 d), x is the exposure concentration (?mol/kg dry soil), a
is the LC50 or the EC50 value, b is the maximal slope of the
logistic curve, and c is the average survival or reproduction
in the solvent control. For the LC50 and EC50 values, the
corresponding pore-water concentrations were calculated us-
ing the carbon content of the LUFA 2.2 soil and averages of
experimentally determined (sorption to humic acids of differ-
ent origin) organic carbon–water partitioning coefficients (Koc)
as reported by Jonassen et al  and by Thomsen . Dif-
ferences between effect concentrations were tested for signif-
icance at the 5% level by comparing the concentration–re-
sponse plots with a generalized likelihood ratio test in Systat
5.2 (Systat, Evanston, IL, USA, 1992).
RESULTS AND DISCUSSION
The extraction procedure applied in the present study re-
sulted in recoveries between 80 and 105% for all tested com-
pounds. Concentrations of anthracene, benz[a]anthracene, py-
rene, benzo[a]pyrene, acridine, acridone, and phenanthridi-
none after 28 d were still at least 70% of the actual initial
concentration. In contrast, naphthalene and quinoline concen-
trations at day 10 had dropped to 7 to 20% of the initial
concentrations. All phenanthrene concentrations were still at
least 70% of the initial concentration at day 10, but in treat-
ments with less than 1,200 ?mol/kg, the concentration dropped
to 5 to 35% in the following period (recovery of treatments
with 550, 842, and 1,459 ?mol/kg was 13, 34, and 96%, re-
spectively, at the end of the enchytraeid test). A similar pattern
was found for tests with phenanthridine (recovery in the 241,
629, and 2,700 ?mol/kg treatments was 30, 60, and 81%,
respectively, at the end of the enchytraeid test).
Average control survival was greater than 80% for both
test species. The average number of juveniles after 28 d in the
control treatments was 683 ? 189 (n ? 6) for F. candida and
983 ? 171 (n ? 8) for E. crypticus. The spiking procedure
with acetone had no effects on survival or reproduction of both
test organisms. The solvent controls were used for further
In the toxicity tests, the effects of the 11 selected com-
pounds on the two endpoints of the two test species were
determined. Figure 1 shows the concentration–response rela-
tionships for phenanthrene, its heterocyclic analogue phen-
anthridine, and its corresponding metabolite phenanthridinone.
Both test species responded differently to the selected PACs,
ranging from clear concentration–response relationships to no
effects at the highest concentrations tested (Fig. 1). The cal-
culated LC10, LC50, EC10, and EC50 values for F. candida
and E. crypticus are presented in Table 2. When two consec-
utive PAC concentrations resulted in no effect and a complete
effect, respectively, no EC50 or LC50 value could be esti-
mated, and the highest no-effect and the lowest total-effect
concentrations are given instead. When the available concen-
tration–response relationships were compared with a gener-
alized likelihood ratio test, effect concentrations with over-
lapping 95% confidence limits in all cases did not differ sig-
nificantly (generalized likelihood ratio test, p ? 0.05), whereas
nonoverlapping 95% confidence limits indicated a significant
difference (generalized likelihood ratio test, p ? 0.05).
The hydrophobic PACs benz[a]anthracene and ben-
zo[a]pyrene as well as anthracene did not affect both soil
invertebrates at the highest tested concentrations (Table 2).
The relatively low maximum water solubility of hydrophobic
PACs in soils probably is a limiting factor for the induction
of effects. With increasing hydrophobicity, the water solubility
of the PACs is decreasing more than the accumulation in the
organism’s lipids is increasing . Pearlman et al.  and
Mackay et al.  reported that the water solubility of an-
thracene is 18- to 20-fold lower than that of phenanthrene,
which explains its lower toxicity. For all three compounds, it
would serve no useful purpose to test higher concentrations,
because it is nearly impossible to ensure a homogeneous PAC
distribution in the soil because of the coagulation of the PAC
while the acetone evaporates during the spiking process.
Comparisons of the effects of naphthalene, anthracene, and
phenanthrene and their azaarene-analoguesquinoline,acridine,
and phenanthridine clearly demonstrate the impact of thesingle
N in-ring substitution on toxicity. However, no clear pattern
could be observed in the changed toxicity: Anthracene was
less toxic than acridine, most likely for the reasons discussed
above. In contrast, naphthalene was more toxic than quinoline,
which matches with its higher log Kow. However, although
phenanthrene has a higher log Kowthan its heterocyclic ana-
logue phenanthridine, they were equally toxic. Hence, the ob-
served differences in toxicity between the homocyclic PACs
and their heterocyclic analogues are not well explained by log
Toxicity of PACs depends on their bioaccumulation from
the pore water into the organisms. The freely dissolved PAC
concentrations in the pore water depend on the soil–pore water
partitioning coefficients. These two parameters are not pro-
portionally related to Kow; therefore, pore-water LC50 and
EC50 values (?mol/L) were calculated using Kocvalues from
Table 1 and the organic carbon content of the LUFA 2.2 soil.
The pore-water LC50 and EC50 values were then again plotted
against log Kow(Fig. 2). For both organisms, the inverse log-
transformed LC50 values showed a positive relationship with
the log Kowof the test compounds. Bleeker et al.  tested
Chronic effects of PACs on soil invertebrates
Environ. Toxicol. Chem. 25, 2006 2427
Fig. 1. Concentration–response relationships for the effects of phenanthrene (PHE), phenanthridine (PHI), and phenanthridinone (PHO) on survival
(black symbols) and reproduction (open symbols) of Folsomia candida (squares) and Enchytraeus crypticus (triangles) after 28 d of exposure
in Landwirtschaftliche Untersuchungs- und Forschungsanstalt (LUFA) 2.2 soil. Symbols represent the means, and error bars indicate minimum
and maximum values. On the x-axis, the actual polycyclic aromatic compound concentrations in the soil at the start of the experiment are plotted.
The curve-fitting procedure was performed as described by Mackay et al. .
a series of PACs on larvae of the midge Chironomus riparius
in a 96-h aquatic toxicity test. In Figure 2, the relationship
between LC50 data and log Kowfrom the acute aquatic toxicity
tests by Bleeker at al.  is presented as a gray line, with
the corresponding 95% confidence limits as dotted lines. The
plot shows that for those PACs that do exert toxicity in the
present study, the effect concentrations for F. candida gen-
erally are well described by the relationship obtained for the
midge C. riparius, whereas E. crypticus clearly is less sen-
sitive than F. candida and C. riparius (see discussion below).
Similar results were found by Sverdrup et al. : Soil–pore
water EC10 values for the springtail F. fimetaria showed a
fairly good correlation with no-observed-effect concentrations
for Daphnia magna. This suggests a narcotic mode of action
for the effects of both homocyclic PACs and azaarenes on
survival of these widely differing invertebrates during chronic
Isomers and metabolites
The present study included three isomer pairs: Anthracene
and phenanthrene, the azaarenes acridine and phenanthridine,
and the metabolites acridone and phenanthridinone. The two
metabolites did not affect both soil invertebrates at the highest
tested concentrations, in contrast to their parent compounds
acridine and phenanthridine. For these two azaarene isomers,
a clear difference in effect concentration was observed, with
phenanthridine being more toxic than acridine. This difference
disappeared, however, when the effects were expressed as
pore-water LC50 values (Fig. 2). Anthracene did not affect
both soil invertebrates at the highest tested concentrations,
whereas for phenanthrene, clear concentration–response rela-
tionships were observed. This difference may be well ex-
plained by the low solubility of anthracene (see above); how-
ever, it is not neutralized by expressing effects as pore-water
LC50 values. Such strong, isomer-specific toxicities have been
observed previously [8–13], and they clearly require refined
molecular modeling to be explained.
Deviations from the relationship between effect concentra-
tions and log Kowvalues may indicate a specific mode of action.
To detect if the selected PAC had a specific effect on repro-
duction, the graphs of the pore-water LC50 and EC50 values
plotted against log Kow(Fig. 2) were compared. Additional
evidence was obtained by calculating LC50 to EC50 ratios for
the PACs that exerted adverse effects (Table 3). The results
indicate a specific effect of phenanthridine on the reproduction
of both species, of phenanthrene on E. crypticus, and espe-
cially of pyrene on the parthenogenetic F. candida. Sverdrup
et al.  also found an effect of pyrene on the sexually
Environ. Toxicol. Chem. 25, 2006S.T.J. Droge et al.
Table 2. Polycyclic aromatic compound (PAC) concentrations in the soil at which 50 and 10% mortality (LC50 and LC10, respectively) and 50 and 10% diminished reproduction (EC50 and EC10,
respectively) occurred for the springtail Folsomia candida and the enchytraeid Enchytraeus crypticus after 28 d of exposure in Landwirtschaftliche Untersuchungs- und Forschungsanstalt (LUFA)
Heterocyclic PACs (azaarenes)
aEffect concentrations are based on measured initial concentrations in ?mol/kg dry soil. Values in parentheses represent the 95% confidence interval.
bWhen two consecutive PAC concentrations resulted in no effect and a complete effect, respectively, then no EC50 or LC50 could be estimated, and the highest no-effect and the lowest total-effect
concentrations are given instead.
cNo reliable effect concentration could be calculated, but data suggest that this test concentration is very close to the LC50 (survival varies between 0 and 80%).
Chronic effects of PACs on soil invertebrates
Environ. Toxicol. Chem. 25, 20062429
Fig. 2. Calculated pore-water concentrations of polycyclic aromatic
compounds (PACs) at 50% effect levels at the start of the 28-d soil
toxicity tests. In the upper graph, the 50% lethal concentrations(LC50
values) for the polycyclic aromatic compounds are plotted against
their Kowvalues; the lower graph shows a plot of the 50% effect
concentrations (EC50 values). Data for Folsomia candida are rep-
resented by squares, and data for Enchytraeus crypticus are repre-
sented by triangles. In both graphs, the solid line and the dotted lines
represent the linear trend between 96-h LC50 data and Kowand the
95% confidence intervals, respectively, obtained from water-only ex-
posure of midge larvae to similar PACs .
Table 3. Ratios between the 50% lethal concentration (LC50) and the
50% effect concentration (EC50) of polycyclic aromatic compounds
(PACs) in Landwirtschaftliche Untersuchungs- und Forschungsanstalt
(LUFA) 2.2. soil for Folsomia candida and Enchytraeus crypticusa
Heterocyclic PACs (azaarenes)
aRatios of 3 or greater, suggesting a specific effect on reproduction,
are given in italic.
reproducing F. fimetaria, demonstrating that the effect of py-
rene is independent of the way the collembolans are repro-
ducing. The specific effect of pyrene on springtail reproduction
may have been caused by toxic metabolites. Hauser et al. 
found that the primary pyrene metabolite, 1-hydroxypyrene,
was both acutely toxic in the Microtox? test (Strategic Di-
agnostics, Newark, DE, USA; 50% inhibition of biolumines-
cence of the luminescent bacteria Vibrio fischeri at 0.68 mg/
L) and genotoxic in the Mutatox test (inducing the ability to
produce light in dark mutants of V. fischeri at 0.313 mg/L).
For the isopod Porcellio scaber and the springtail Orchesella
cincta, the primary metabolite 1-hydroxypyrene and the phase-
2 metabolites pyrene-1-glucoside and pyrene-1-conjugatewere
measured rapidly after the start of oral exposure or exposure
at contaminated field sites [31,32]. The rate of PAC metabo-
lism in isopods and springtails contrasts with that in earth-
worms, which metabolize PACs relatively slowly . Al-
though no direct data are available regarding the biotransfor-
mation capacity of enchytraeids, toxic metabolites likely
would not have been formed to the same extent as in the
springtails. This could explain the finding that pyrene was
much less toxic to the enchytraeids than to the springtails. A
further agreement between the present study and that by Sver-
drup et al.  is the effect of phenanthrene on enchytraeid
reproduction, but remarkable differences also were observed:
In the study by Sverdrup et al. , pyrene also affected en-
chytraeid reproduction, but this was not the case in the present
study. It is concluded that specific effects on reproduction,
which become evident during chronic exposure, are compound
and species specific. The unpredictability of these effects ques-
tions the applicability of quantitative structure–activity rela-
tionships and acute to chronic ratios for predicting chronic
sublethal effects of PACs.
For all homocyclic PACs and azaarenes that did affect re-
production and survival at a certain level, F. candida appeared
to be more sensitive than E. crypticus. Remarkable differences
are the 32-fold lower EC50 value for springtails after pyrene
exposure in soil and the 6.5-fold lower LC50 value after quin-
oline exposure (Table 2). Pyrene, phenanthrene, and acridine
had been tested previously on a related collembolan (F. fi-
metaria) and the same enchytraeid [15–17], and also in those
studies, the collembolan was more sensitive than the enchy-
For F. fimetaria, a larger data set is available that shares
six compounds with the present study. Generally, the results
of both studies are in agreement, with benz[a]anthracene and
benzo[a]pyrene not being toxic at the highest test concentra-
tions and acridine being less toxic than the homocyclic PACs.
Yet, a very distinct difference for anthracene was observed:
In the present study, anthracene was not toxic at all, but it was
more than sixfold more toxic to F. fimetaria than predicted
from its log Kow.
Comparisons between the few studies of terrestrial inver-
tebrates are complicated because of the wide variety of test
conditions and experimental setups. Still, some general trends
may be derived: Sverdrup et al.  demonstrated that for the
majority of PACs (eight compounds tested, sharing three with
the present study), springtails were more sensitive than oli-
gochaetes (earthworms and enchytraeids). The lower sensitiv-
ity of enchytraeids compared to springtails also was observed
in the present study. The present F. candida EC50 value for
phenanthrene obtained in a soil with 3.9% organic matter is
in the same range as that for F. fimetaria in soil with 2.8%
organic matter , but this value is much lower than the
effect concentrations reported in soils with 10% organic matter
[34,35], most likely because of a lower availability in the soils
with the higher organic matter content. Both Sverdrup et al.
Environ. Toxicol. Chem. 25, 2006 S.T.J. Droge et al.
[15–18] and the present study found no effects of PACs with
a log Kowgreater than 5.2. In contrast, Bauer and Pohl 
reported that reproduction of F. candida already was affected
at a benzo[a]pyrene concentration of 10 mg/kg, and effects of
benzo[a]pyrene have been described for enchytraeids  and
the isopod P. scaber  as well. It is concluded that although
some generalizations can be made, widely differing effect con-
centrations also have been reported, which can be explained
only in part by differences in experimental setups.
The results of the present study demonstrate that as far as
narcosis-induced mortality is concerned, effects of both ho-
mocyclic and heterocyclic PACs are well described by the
relationship between pore-water LC50 values and log Kow. In
contrast, specific effects on reproduction varied between spe-
cies and between compounds as closely related as isomers.
These unpredictable specific effects on reproduction showed
up as deviations from the relationship between pore-water
EC50 values and log Kow, and they force one to test the toxicity
of these PACs to populations of soil invertebrates to obtain
reliable effect concentrations for use in risk assessment of
PACs. The long-term consequences of exposure to PACs in
the field, however, should be analyzed in multigeneration ex-
Acknowledgement—This research was supported by the Ministry of
Housing, Spatial Planning and the Environment (Martine van der
Weiden), and by the Technology Foundation (STW), applied science
division of The Netherlands Organization for Scientific Research
(NWO), and the technology program of the Ministry of Economic
Affairs (project AEB 6364). We thank Martien Janssen, Eric Ver-
bruggen (National Institute for Public Health and the Environment
[RIVM]), and Tom Parkerton and coworkers (ExxonMobil) for their
comments on the manuscript.
1. Nielsen T, Clausen P, Jensen FP. 1986. Determination of basic
azaarenes and polynuclear aromatic hydrocarbons in airbornepar-
ticulate matter by gas chromatography. Anal Chim Acta 187:223–
2. Nielsen T, Feilberg A, Binderup ML, Tønnesen J. 1999. Impact
of regulations of traffic emissions on PAH level in the air. En-
vironmental Project Report 447. Danish EnvironmentalProtection
Agency, Copenhagen, Denmark.
3. Michael JP. 2000. Quinoline, quinazoline and acridone alkaloids.
Natural Product Reports 17:603–620.
4. Oshiro Y, Sato S, Kurahashi N, Tanaka T, Kikuchi T, Tottori K,
Uwahodo Y, Nishi T. 1998. Novel antipsychotic agents with do-
pamine autoreceptor agonist properties: Synthesis and pharma-
cology of 7-[4-(4-phenyl-1- piperazinyl)butoxy]-3,4-dihydro-
2(1H)-quinoline derivatives. J Med Chem 41:658–667.
5. Siim BG, Hicks KO, Pullen SM, van Zijl PL, Denny WA, Wilson
WR. 2000. Comparison of aromatic and tertiary amine N-oxides
of acridine DNA intercalators as bioreductive drugs—Cytotox-
icity, DNA binding, cellular uptake, and metabolism. Biochem
6. Kuhn EP, Suflita JM. 1989. Microbial degradation of nitrogen,
oxygen, and sulfur heterocyclic compounds under anaerobic con-
ditions. Environ Toxicol Chem 8:1149–1158.
7. Crommentuijn T, Sijm D, de Bruin J, van Leeuwen K, van de
Plassche E. 2000. Maximum permissible and negligible concen-
trations for some organic substances and pesticides. J Environ
8. Walton BT, Ho CH, Ma CY, O’Neill EG, Kao GL. 1983. Ben-
zoquinolinediones: Activity as insect teratogens. Science 222:
9. Wood AW, Chang RL, Levin W, Ryan DE, Thomas PE, Lehr RE,
Kumar S, Schaefer-Ridder M, Engelhart U, Yagi H, Jerina DM,
Conney AH. 1983. Mutagenicity of diol-epoxides and tetrahy-
droepoxides of benz[a]acridine and benz[c]acridine in bacteria
and in mammalian cells. Cancer Res 43:1656–1662.
10. Kumar S, Sikka HC, Dubey SK, Czech A, Geddie N, Wange CX,
LaVoie EJ. 1989. Mutagenicity and tumorigenicity of dihydro-
diols, diol epoxides, and other derivatives of benzo[f]quinoline
and benzo[h]quinoline. Cancer Res 49:20–24.
11. Kraak MHS, Wijnands P, Govers HAJ, Admiraal W, de Voogt P.
1997. Structural-based differences in ecotoxicity of benzoquin-
oline isomers to the zebra mussel (Dreissena polymorpha). En-
viron Toxicol Chem 16:2158–2163.
12. Bleeker EAJ, van der Geest HG, Kraak MHS, de Voogt P, Ad-
miraal W. 1998. Comparative ecotoxicity of NPAHs to larvae of
the midge Chironomus riparius. Aquat Toxicol 41:51–62.
13. Wiegman S, van Vlaardingen PLA, Bleeker EAJ, de Voogt P,
Kraak MHS. 2001. Phototoxicity of azaarene isomers to the ma-
rine flagellate Dunaliella tertiolecta. Environ Toxicol Chem 20:
14. Achazi RK, van Gestel CAM. 2003. Uptake and accumulation of
PAHs by terrestrial invertebrates. In Douben PET, ed, PAHs: An
Ecotoxicological Perspective. John Wiley, Chichester, UK, pp
15. Sverdrup LE, Kelley AE, Krogh PH, Nielsen T, Jensen J, Scott-
Fordsmand JJ, Stenersen J. 2001. Effects of eight polycyclic ar-
omatic compounds on the survival and reproduction of the spring-
tail Folsomia fimetaria L. (Collembola, Isotomidae). Environ
Toxicol Chem 20:1332–1338.
16. Sverdrup LE, Jensen J, Kelley AE, Krogh PH, Stenersen J. 2002.
Effects of eight polycyclic aromatic compounds on the survival
and reproduction of Enchytraeus crypticus (Oligochaeta, Clitel-
lata). Environ Toxicol Chem 21:109–114.
17. Sverdrup LE, Krogh PH, Nielsen T, Stenersen J. 2002. Relative
sensitivity of three terrestrial invertebrate tests to polycyclic ar-
omatic compounds. Environ Toxicol Chem 21:1927–1933.
18. Sverdrup LE, Nielsen T, Krogh PH. 2002. Soil ecotoxicity of
polycyclic aromatic hydrocarbons in relation to soil sorption, li-
pophilicity, and water solubility. Environ Sci Technol 36:2429–
19. International Organization for Standardization. 1999. Soil qual-
ity—Inhibition of reproduction of Collembola (Folsomia can-
dida) by soil pollutants. Guideline 11267. Geneva, Switzerland.
20. Ro ¨mbke J, Moser T. 2002. Validating the enchytraeid reproduc-
tion test: Organization and results of an international ringtest.
21. Organization for Economic Cooperation and Development. 2000.
Enchytraeidae reproduction test. Draft test guideline 220. Paris,
22. Wiles JA, Krogh PH. 1998. Testing with the collembolans Iso-
toma viridis, Folsomia candida, and Folsomia fimetaria. In
Løkke H, van Gestel CAM, eds, Handbook of Soil Invertebrate
Toxicity Tests. John Wiley, Chichester, UK, pp 131–156.
23. Didden W, Ro ¨mbke J. 2001. Enchytraeids as indicator organisms
for chemical stress in terrestrial ecosystems. Ecotoxicol Environ
24. Haanstra L, Doelman P, Oude Voshaar JH. 1985. The use of
sigmoidal concentration–response curves in soil ecotoxicological
research. Plant Soil 84:293–297.
25. Jonassen KEN, Nielsen T, Hansen PE. 2003. The application of
high-performance liquid chromatography humic acid columns in
determination of Kocof polycyclic aromatic compounds. Environ
Toxicol Chem 22:741–745.
26. Thomsen M. 2002. QSARs in environmentalriskassessment.PhD
thesis. Roskilde University, Roskilde, Denmark.
27. Pearlman RS, Yalkowsky SH, Banerjee S. 1984. Water solubility
of polynuclear aromatic and heteroaromatic compounds. J Phys
Chem Ref Data 13:555–562.
28. Mackay D, Shiu WY, Ma KC. 1999. Physical-Chemical Prop-
erties and Environmental Fate. Chapman & Hall, New York, NY,
29. Bleeker EAJ, Pieters BJ, Wiegman S, Kraak MHS. 2002. Com-
parative (photo-enhanced) toxicity of homocyclic and heterocy-
clic PACs. Polycyclic Aromatic Compounds 22:601–610.
30. Hauser B, Schrader G, Bahadir M. 1997. Comparison of acute
toxicity and genotoxic concentrations of single compounds and
waste elutriates using the Microtox/Mutatox test system. Ecotox-
icol Environ Saf 38:227–231.
31. Howsam M, Van Straalen NM. 2004. Pyrene metabolism in the
Chronic effects of PACs on soil invertebrates Download full-text
Environ. Toxicol. Chem. 25, 2006 2431
springtail Orchesella cincta L. (Collembola, Entomobryidae).
Environ Toxicol Chem 22:1481–1486.
32. Stroomberg GJ, Ariese F, Van Gestel CAM, Van Hattum B, Vel-
thorst NH, Van Straalen NM. 2004. Pyrene biotransformationand
kinetics in the hepatopancreas of the isopod Porcellio scaber.
Arch Environ Contam Toxicol 47:324–333.
33. Jager T, Sanchez FAA, Muijs B, van der Velde EG, Posthuma L.
2000. Toxicokinetics of polycyclic aromatic hydrocarbons in Ei-
senia andrei (Oligochaeta) using spiked soil. Environ Toxicol
34. Bowmer CT, Roza P, Henzen L, Degeling C. 1992. The devel-
opment of chronic toxicological tests for PAH contaminated soils
using the earthworm Eisenia fetida and the springtail Folsomia
candida. TNO Report IMW-R 92/387. Apeldoorn, The Nether-
35. Crouau Y, Chenon P, Gisclard C. 1999. The use of Folsomia
candida (Collembola, Isotomidae) for the bioassay of xenobiotic
substances and soil pollutants. Appl Soil Ecol 12:103–111.
36. Bauer H, Pohl D. 1998. Bodeno ¨kologische Untersuchungen zur
Wirkung und Verteilung von organischen Stoffgruppen (PAK,
PCB) in ballungsraumtypischen O¨kosystemen.Forschungsbericht
des Projekttra ¨gers 1/98, 165p. GSF-Forschungszentrum fu ¨r Um-
welt und Gesundheit, Neuherberg, Germany.
37. Achazi RK, Chroszcz G, Du ¨ker C, Henneken M, Rothe B, Schaub
K, Steudel I. 1995. The effect of fluoranthene (Fla), ben-
zo[a]pyrene (BaP) and cadmium (Cd) upon survival rate and life
cycle parameters of two terrestrial annelids in laboratory test
systems. Newsletter on Enchytraeidae 4:7–14.
38. Van Brummelen TC, van Gestel CAM, Verweij RA. 1996. Long-
term toxicity of five polycyclic aromatic hydrocarbons for the
terrestrial isopods Oniscus asellus and Porcellio scaber. Environ
Toxicol Chem 15:1199–1210.
39. Helweg C, Nielsen T, Hansen PE. 1997. Determination ofoctanol-
water partition coefficients of polar polycyclic aromatic com-
pounds (N-PAC) by high performance liquid chromatography.
40. Bleeker EAJ. 1999. Toxicity of azaarenes: Mechanisms and me-
tabolism. PhD thesis. University of Amsterdam, Amsterdam, The