Environ. Sci. Technol. 2002, 36, 2429-2435
Soil Ecotoxicity of Polycyclic Aromatic Hydrocarbons in Relation to Soil Sorption, Lipophilicity, and Water Solubility L I N E E . S V E R D R U P , * ,†,‡ TORBEN NIELSEN,§ AND PAUL HENNING KROGH| JordforsksCentre for Soil and Environmental Research, N-1432 Ås, Norway, University of Oslo, Department of Biology, P.O. Box 1050 Blindern, N-0316 Oslo, Norway, Risø National Laboratory, PBK 313, P.O. Box 49, DK-4000 Roskilde, Denmark, and National Environmental Research Institute, Vejlsøvej 25, DK-8600 Silkeborg, Denmark
A data set was generated aiming to predict the toxicity of PAHs to soil organisms. Toxicity data include the effects of 16 PAHs on the survival and reproduction of the soil-dwelling springtail Folsomia fimetaria. The results show that only PAHs with reported log Kow values e 5.2 (i.e., naphthalene, acenaphthene, acenaphthylene, anthracene, phenanthrene, fluorene, pyrene, and fluoranthene) significantly affected the survival or reproduction of the test organisms. Threshold values for the toxicity of the individual PAHs could be expressed as pore-water concentrations by the use of reported organic carbon-normalized soil-pore-water partitioning coefficients (Koc values). For the PAHs with a log Kow e 5.2, toxicity significantly increased with increasing lipophilicity of the substances (r2 ) 0.67; p ) 0.012; n ) 8), suggesting a narcotic mode of toxic action for most substances. However, the position of anthracene in the regression plot indicated a more specific mode of toxic action than narcosis, and removing this data point yielded the following regression equation: log EC10 (µmol/L) ) -0.97 log Kow + 4.0 (r2 ) 0.80; p ) 0.006; n ) 7). Using this quantitative structure-activity relationship (QSAR) to calculate threshold values for the toxicity of the remaining nontoxic substances (benz[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, dibenz[a,h]anthracene, benzo[a]pyrene, perylene, and indeno[1,2,3-cd]pyrene), the absence of toxicity could, in most cases, be explained by a limited water solubility, indicating that these substances do act by narcosis as the mode of toxic action and that their toxicity is governed by concentrations in the pore-water. The results provide important input to future model predictions of the ecological risk posed by PAH contaminated sites. * Corresponding author phone: +47 64 94 81 63; fax: +47 64 94 81 10; e-mail:
[email protected]. Present address: Norwegian Centre for Soil and Environmental Research, Frederik A. Dahls vei 20 N-1432 Ås, Norway. † Jordforsk. ‡ University of Oslo. § Risø National Laboratory. | National Environmental Research Institute. 10.1021/es010180s CCC: $22.00 Published on Web 04/24/2002
2002 American Chemical Society
Introduction Contamination by polycyclic aromatic hydrocarbons (PAHs) is one of the main problems related to industrial soils. Still, for soil organisms, the effects of the individual constituents of these mixture pollutants are not properly investigated. The complex nature of PAH mixtures in soils makes it unlikely that ecotoxicity data for each individual compound will ever be generated. However, the use of physical-chemical properties to predict toxicity of substances other than those tested (i.e., quantitative structure-activity relationships (QSARs)) may be a solution to this problem. Because of the lack of functional groups, PAHs have been expected to act by narcosis as the mode of toxic action. For soil organisms, a recent study with springtails supports this assumption (1). Narcotic chemicals do not bind to specific receptors within an organism but elicits toxicity by affecting the fluidity and function of cell membranes. Already in 1899, Meyer (2) proposed to use fat-water partitioning coefficients to explain the difference in narcotic activity of many substances, and it is now well-known that the toxicity of narcotic chemicals to aquatic organisms increases with increasing lipopilicity (3). To derive a QSAR for the toxicity of PAHs to soil organisms, however, the bioavailability of the chemicals must also be taken into consideration. Several studies by van Gestel and co-workers (4-6) demonstrated that the toxicity of five chlorophenenols, 1,2,3-trichlorobenzene, and 2,4-dichloroaniline to earthworms is primarily governed by the concentration in the pore-water. The idea that toxicity is governed by the concentration in pore-water is supported by some studies (7, 8), but there are others suggesting that, for highly lipophilic substances, additional uptake through the gut might be important (9, 10). Pore-water concentrations for neutral organic substances (in freshly spiked soils) can be calculated from total soil concentrations by the use of soil organic carbon-pore-water partitioning coefficients (Koc values). Hence, if soil organisms are mainly exposed through pore-water and all PAHs have a narcotic mode of toxic action, the toxicity of individual PAHs to soil organisms can be described by their lipophilicity (log Kow) and soil sorption properties (i.e., log Koc combined with information on the organic carbon content of the soil used). This is in line with the three-phase model involving soil-to-pore-water and pore-water-to-earthworm partitioning proposed by van Gestel et al. (6). However, some highly hydrophobic chemicals, for which a very high toxicity can be predicted, are not acutely or chronically toxic to aquatic organisms. In 1981, Ko¨nemann (11) proposed that the low water solubility, in combination with a limited bioconcentration potential, could explain the absence of toxicity for such compounds to fish. For soil organisms, the question arises whether these organisms are exposed only through pore-water or if, for example, ingestion of soil particles may significantly increase the uptake of hydrophobic contaminants. Whether highly lipophilic substances are toxic to soil organisms or not is therefore an important input to realistic model predictions on PAH ecotoxicity. To obtain a high quality dataset for the prediction of PAH ecotoxicity and to test the hypothesis that water solubility also limits the toxicity of some highly lipophilic PAHs to soil organisms, 16 PAHs, with n-octanol-water partitioning coefficients (log Kow) ranging from 3.3 to 6.7 were selected for toxicity testing. Twelve of these chemicals were tested as part of this study, while four others have earlier been tested VOL. 36, NO. 11, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2429
TABLE 1. Molecular Weight (MW), Water Solubility (WS), n-Octanol-Water Partitioning Coefficient (log Kow), Soil Organic Carbon-Water Partitioning Coefficient (log Koc), and Estimated Soil-Water Partitioning Coefficient (Kd) in the Askov Soil for 16 PAHs substance
MW
naphthalene acenaphthylene acenaphthene fluorene anthracene phenanthrene pyrene fluoranthene benz[a]anthracene chrysene benzo[b]fluoranthene benzo[k]fluoranthene perylene benzo[a]pyrene indeno[1,2,3-cd]pyrene dibenz[a,h]anthracene
128.2 152.2 154.2 166.2 178.2 178.2 202.2 202.2 228.3 228.3 252.3 252.3 252.3 252.3 276.3 278.4
WS (µg/L) log Kow 31700 (22) 3930 (23) 16100 (22) 1980 (22) 73 (22) 1290 (22) 135 (22) 260 (22) 14 (22) 2.0 (22) 1.5 (24) 0.8 (24) 0.4 (22) 3.8 (22) 0.19 (25) 0.6 (26)
3.32 (27) 4.07 (28) 3.94 (27) 4.23 (27) 4.50 (29) 4.60 (29) 5.20 (29) 5.20 (29) 5.66 (29) 5.80 (29) 6.40 (29) 6.40 (29) 6.40 (29) 6.20 (29) 6.70 (30) 6.50 (31)
log Koc
Kda
3.11 (32) 3.83 (33) 3.79 (33) 4.15 (33) 4.41 (33) 4.22 (33) 4.82 (33) 4.74 (33) 5.252 5.372 5.89b 5.89b 5.89b 5.71b 6.14b 5.97b
21 108 99 226 542 266 1060 879 2840 3750 12300 12300 12300 8280 22300 15000
a K ) K × 0.016 (fraction of organic carbon). b Estimated from K d oc ow using the regression formula obtained from Figure 1; log Koc ) 0.8613 log Kow + 0.374.
FIGURE 1. Regression of soil sorption data (log Koc values) on the n-octanol-water partitioning coefficients (log Kow values) for eight polycyclic aromatic hydrocarbons. in the same soil type (1). The experiments were carried out using the small, common springtail species Folsomia fimetaria as test organism.
Materials and Methods Test Substances. Naphthalene (99%), acenaphthene (99%), acenaphthylene (analyzed, 87%), anthracene (99%), chrysene (>95%), benz[a]anthracene (99%), benzo[k]fluoranthene (98%), perylene (99%), benzo[b]fluoranthene, benzo[a]pyrene (>98%), dibenz[a,h]anthracene (97%), and indeno[1,2,3-cd]pyrene was purchased from Sigma-Aldrich (St. Louis, MO). Acetone (J.T. Baker, Phillipsburg, NJ; HPLC quality) was used as spiking solvent. Selected physical-chemical data on these 12 substances and four others (fluorene, phenanthrene, pyrene, and fluoranthene), earlier tested for toxicity (1), are included in Table 1. To characterize bioavailability, data on soil sorption (log Koc) was needed. However, for many of the higher molecular weight PAHs (i.e., those with log Kow g 5.7), literature values were absent or varied enormously between studies (12). To obtain a reasonably homogeneous data set for soil sorption, Koc values for these substances were estimated from their log Kow, based on the relationship found between the log Kow and the log Koc for the PAHs with log Kow e 5.2 (Figure 1). 2430
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 11, 2002
Sample Preparation. An agricultural soil (Askov soil) with an organic carbon content of 1.6% was used. Samples were sieved through a 2 mm mesh prior to use. The Askov soil is a sandy loam, and it has the following particle size distribution: coarse sand (200-2000 µm) 38.4%, fine sand (63-200 µm) 23.6%, coarse silt (20-63 µm) 10.0%, fine silt (2-20 µm) 12.3%, and clay (980 >1030 >360 >560 >560 >840 >910 >780
63 8.8 7.0 1.0 0.69 0.87 0.25 0.46
EC10 valuesa mg/kgb µmol/Lc 20 (0, 39) 23 (12, 31) 31 (0, 42) 7.7 (5.2, 10) 5 (3.2, 12) 23 (9.1, 38) 10 (7.3, 13) 37 (26, 48) >980 >1030 >360 >560 >560 >840 >910 >780
7.6 1.4 2.0 0.21 0.052 0.49 0.047 0.21
QSAR-estimated EC10 values (µmol/L)d
expected EC10 < solubility?
6.0 1.5 1.1 0.79 0.43 0.35 0.090 0.090 0.032 0.024 0.0062 0.0062 0.0062 0.0097 0.0032 0.0050
yes yes yes yes no yes yes yes yes no no no no yes/no g no no
a For the mg/kg data, the 95% confidence intervals are given. b Determined on the basis of measured initial exposure concentrations. c Calculated using the soil-pore-water partitioning coefficients and molecular weights in Table 1. d Quantitative structure-activity relationship between log Kow and log EC10 (µmol/L pore-water) for the eight substances that elicited toxicity: log EC10 (µmol/L) ) -0.97 log Kow + 4.0. e ne ) could not be estimated. f Data from ref 1. g Reported solubility values are both above and below the estimated threshold value for toxicity (12).
TABLE 3. Results of the Chemical Analysis
substance
measured concentrations (mg/kg dry wt) for the 1000, 100, and 10 mg/kg sampling measured/ exposure levels time (t) nominal (days) 1000 100 10 values (%)
naphthalene naphthalene acenaphthene acenaphthene acenaphthylene acenaphthylene anthracene anthracene benz[a]anthracene benzo[b]fluoranthene benzo[a]pyrene dibenz[a,h]anthracene chrysene benzo[k]fluoranthene perylene indeno[1,2,3-cd]pyrene
0 21 0 21 0 21 0 21 0 0 0 0 0 0 0 0
780 29 1000 490
840 840 980 360 840 780 1030 560 560 910
74 3.0 98 23 150 43 79 66
8.1 0.79 10 2.8 16 6.0 8.3 5.8
77.5 4.6 101.7 32.9 156.3 51.3 81.7 69.2
(Merck, p.a.). Each extraction lasted 30 min. The combined extracts were concentrated to 1.5 mL. Identification and quantification of compounds were carried out with a gas chromatography-mass spectrometric system consisting of a Varian 3400 Star GC and a Varian Saturn III ion trap MS, using temperature programmable splitless injection and a 30-m XTI-5 fused silica capillary column coated with 95% methylsiloxane/5% phenylsiloxane (Restek Corp., Bellefonte, PA). In the quantification of the compounds, PAC concentrations were corrected for recovery of the internal standards. In that way, losses during the extraction were corrected for. In test experiments with pyrene added to pure soil, recoveries of pyrene ranged from 95% to 100%. Results of the chemical analysis are given in Table 3.
Results and Discussion PAH Toxicity. Eight out of 16 substances affected the survival or reproduction of F. fimetaria within the concentration range tested (Table 2). The estimated threshold value for toxicity (10% reduction in reproductive output - EC10 value) and
estimated 50% lethal concentration (LC50 value) for these substances are given in Table 2. From the results presented, it is evident that only the PAHs that have a low to intermediate lipophilicity (log Kow in the range 3.3-5.2) are toxic to F. fimetaria. Some typical dose-response relationships are given in Figure 2. QSAR for PAH Toxicity. On the basis of a subset of the eight substances that elicited toxicity, the strategy for data analysis was to perform a regression of the toxicity (on the basis of molar soil pore-water concentrations) on the lipophilicity (log Kow) of the substances in question. The test results (LC50 and EC10 values), expressed as molar pore-water concentrations (see Table 2), show significantly negative relationships with the lipophilicity of the substances (Figure 3A). A better linear fit of the toxicity on log Kow is obtained using the survival data (LC50 values: r2 ) 0.88; p ) 0.0005; n ) 8) than using the reproduction data (EC10 values: r2 ) 0.67; p ) 0.024; n ) 8). However, much of the unexplained variation in the regression with the EC10 values is caused by the substance anthracene being more than 6 times more toxic than predicted by its log Kow. The position of anthracene in the plot of EC10 values on the log Kow of the substances suggests that it may have a more specific mode of toxic action than narcosis. This was also suggested by Kalf et al. (19), who compared QSAR estimated threshold values to experimental data for algae and daphnids and found that anthracene was more than 10 times more toxic to this organisms than what was predicted by its log Kow. Following removal of the data point for anthracene, an r2 value of 0.80 (p ) 0.006; n ) 7) was obtained (Figure 3B), but the regression equation changed only slightly (Figure 3B vs 3A). Still, on the basis that the QSAR should be based only on narcotic substances, the regression equation where anthracene was excluded was used for further calculations. Water Solubility versus Toxicity for Highly Lipophilic PAHs. The water solubility of PAHs shows a strong negative relationship with the log Kow (Figure 4). To check if limitations in water solubility were likely to be the reason that the highly lipophilic PAHs were not toxic in this study, the relationship obtained between lipophilicity and toxicity (Figure 3B) was used to predict toxic concentrations for these eight substances. By comparing the QSAR estimated toxic threshold values (Table 2) with reported data on water solubility (Table 1), it can be concluded that six out of the eight chemicals VOL. 36, NO. 11, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2431
FIGURE 2. Dose-response relationship (nominal values) for the effects of naphthalene, acenaphthene, acenaphthylene, anthracene, and benzo[a]pyrene on the survival and reproduction of the springtail Folsomia fimetaria.
FIGURE 3. Regression of toxic concentrations on the n-octanol-water partitioning coefficients (log Kow values) for eight polycyclic aromatic hydrocarbons. LC50 values describe the concentrations estimated to give 50% reduction in survival. EC10 values describe the concentrations estimated to give 10% reduction in reproductive output: (A) EC10 values for all substances included; (B) EC10 value for anthracene excluded. with a log Kow > 5.2 were not expected to be toxic due to limitations in solubility (Table 2). For benzo[a]pyrene, which was not toxic to springtails, water solubility is reported corresponding to both higher and lower values than the toxic threshold concentration (12). The water solubility of benz[a]anthracene should be high enough to cause toxicity 2432
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 11, 2002
according to the model predictions, but no effect could be seen in the toxicity tests. Anthracene has very low solubility as compared to its log Kow value (Table 1), and on the basis of the model predictions, anthracene should not be toxic to springtails (Table 2). Still, a high toxicity was observed for anthracene in the springtail
FIGURE 4. Regression of water solubility data on the n-octanolwater partitioning coefficients (log Kow values) for 16 polycyclic aromatic hydrocarbons.
FIGURE 5. Combined plot with predicted EC10 and LC50 values and predicted water solubility of PAHs in relation to their n-octanolwater partitioning coefficients. The QSAR models used to predict toxicity and solubility have been obtained from Figures 3B and 4, respectively. test, and this illustrates an important drawback of the proposed model: it will underestimate toxicity for substances that act by a more specific mode of toxic action than narcosis. Anthracene thus serves as an example suggesting that PAHs may have more specific modes of toxic action, but the absence of toxicity found for the highly lipophilic substances suggests that these substances do actually act by narcosis as the mode of toxic action. The relationship between toxicity and solubility can also be investigated on a more general basis, using the QSAR model predictions. When the estimated toxicity data (from Figure 3B) and estimated water solubility data (from Figure 4) were plotted against substance lipophilicity in the same figure (Figure 5), it was evident that water solubility, in general, decreases more than the increase in toxicity. From the figure, it can roughly be observed that individual PAHs with a log Kow > 5.5 are not expected to give effects on springtail survival while PAHs with log Kow > 6.0 will no longer affect either survival or reproduction of this organism. Alternative Explanations. Some alternative explanations for the observation that highly lipophilic PAHs are not toxic to springtails can be suggested. For example, the relatively short exposure period (3 weeks) might not be sufficient to achieve equilibrium between pore-water and the organisms. Uptake kinetics of PAHs has not been studied in springtails, but a study by Krauss et al. (20) showed that uptake of PAHs and PCBs in earthworms (Lumbricus terrestris) was near steady-state after about 15 days of exposure. The body mass
of earthworms is much higher than that of springtails, and a steady-state period longer than 15 days for springtails is therefore rather unlikely. On the other hand, uptake of highly lipophilic substances might be counterbalanced by biotransformation, leading to lower accumulation factors for these substances. However, to the knowledge of the authors, there are no studies available suggesting a higher biotransformation rate for highly lipophilic PAHs as compared to those with a lower lipophilicity. Other Studies on Terrestrial Organisms. Studies on the toxicity of highly lipophilic PAHs to other terrestrial organisms could shed light on the validity of the model predictions to organisms other than that tested, but there are not many data available. Bowmer et al. (34) investigated the toxicity of chrysene to the earthworm Eisenia fetida and found no effects on earthworm survival. In 1996, Van Brummelen et al. (35) reported on the effects of benz[a]anthracene and benzo[a]pyrene on the growth of the two isopods, Oniscus asellus and Porcellio scaber, exposed through the food. For O. asellus, small but significant effects on growth was observed after exposure to benz[a]anthracene, and no effects were found after exposure to benzo[a]pyrene. For P. scaber, no effects were observed for any of these substances. From our proposed model, benz[a]anthracene is expected to be toxic while chrysene and benzo[a]pyrene are not. However, there are too few data available to make any conclusions on the validity of the results for organisms other than that tested. Making the Model: Aspects of Uncertainty. The model predictions of toxicity depend strongly on which log Kow and log Koc values are selected for each individual substance. In the literature, several values are available for most of the PAHs used in the present study (see ref 12), and some data had to be selected over others. The use of calculated log Koc values for the highly lipophilic PAHs introduces additional uncertainty for these substances, and the fact that some authors have suggested that log Kow values are not strong predictors of soil sorption for highly lipophilic substances (e.g., ref 21) makes the reliability of these data questionable. On the other hand, we wanted to use a data set that would give the best possible impression of the difference in sorption coefficients between PAHs, and this criterion was most likely to be fulfilled by selecting values from a single study or at least values generated using very similar methods. This was the rationale for selecting log Koc values from only two studies (by the same authors, refs 32 and 33) and using these values to calculate sorption coefficients for the remaining substances that had not been investigated by these authors. Hence, even though we emphasize that it is likely that another set of physicochemical data would give (slightly) different results in the toxicity estimations, the data that were used here has the advantage that the uncertainty usually associated with values obtained from different studies is eliminated. Another factor of uncertainty lies in detecting effects within toxicity tests. The variation between replicate treatments is sometimes substantial (see Figure 2), and this variation will often mask toxicity if the effect obtained at the highest treatment level is not very large (e.g., a maximum of 15% reduction in reproduction relative to control values). Relevance of the Test System. The organism used in the present study, the springtail F. fimetaria, is generally more sensitive to PAHs and some N-, S-, and O-substituted analogues than enchytraeids (36), earthworms (37), soilnitrifying bacteria (38), and three species of terrestrial plants (39) tested in the same soil type. Springtails therefore seem like a good model for the sensitivity of the soil ecosystem to this group of contaminants. All the aforementioned studies (36-39) suggested that toxicity to the test organisms was governed mainly by the pore-water concentrations of the test substances. However, there may be species in the terrestrial ecosystem that are exceptions to this trend, that VOL. 36, NO. 11, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2433
is, whose feeding biology makes additional uptake from soil particles likely. In a study using the marine polychaete Capitella sp., Forbes et al. (40) predicted that, for this benthic deposit-feeding species, fluoranthene dietary uptake flux was 20-30 times higher than that due to uptake from pore-water. Although this study does not give answers to whether this organism is still in equilibrium with the surrounding porewater concentrations or not, it emphasizes the importance of considering feeding biology in the assessment of ecological effects from chemical exposure. Possible Practical Applications. For some individual PAHs, soil quality criteria have been developed (19). However, PAH contaminated soils are typically characterized by very complex mixtures of chemicals. Mixture toxicity for chemicals with a narcotic mode of toxic action can be considered additive (on an internal molar basis), and each substance present in a mixture may therefore contribute to the overall toxicity of a contaminated soil sample. The use of soil quality criteria for individual PAHs are therefore of little relevance to the assessment of the environmental hazard and risk posed by PAH contaminated soils. In 1995, Swartz et al. (41) developed a model (ΣPAH) to predict the summed toxicity of PAH mixtures in sediments. The model was based on a combination of equilibrium partitioning, QSAR, toxic units, and additivity. As pointed out by the authors, the toxicity input data to this model was very limited, as only three substances were tested (41). Still, the model showed a good predictability when model estimates were compared to test results obtained by testing contaminated sediment samples (41). In a study by Ozretich et al. (42), ΣPAH model predictions were further improved by excluding the highly lipophilic PAHs (log Kow > 6.0) from the calculations. Even if developed for sediments, it is obvious to think that the principles of the ΣPAH model can also be applied to soils. The model consists of two main elements: characterization of exposure (in the ΣPAH model, equilibrium partitioning has been used) and characterization of effects (in the ΣPAH model, lethality has been used as the toxic endpoint). The results of the present study support the assumption toxicity is governed by pore-water concentrations; hence, the exposure assessment can be performed assuming equilibrium partitioning by the use of soil-porewater partitioning coefficients (Kd or Koc) values. Because of factors such as aging processes (e.g., ref 43), the characterization of exposure could possibly be improved using nonexhaustive extractions with mild solvents to assess bioavailable fractions (44, 45), even if not all studies show improved predictability of bioavailability using such extraction methods (e.g., ref 20). The new ecotoxicity data presented here is a necessary step forward in the development of ecotoxicity models for PAH contaminated soils. The toxicity of individual PAHs can be assessed by transforming pore-water exposure concentrations (ExC) into toxic units (TU) by comparison with toxic threshold concentrations (TTC): TU ) ExC/TTC. For PAHcontaminated soil samples, expected toxicity of a mixture then equals the sum of TU for all substances present. We are, however, not sure whether our data support the suggestion of Ozretich et al. (42) that PAHs with high log Kow values (e.g., >6.0) should be excluded from calculations from the calculation of toxic units; even if such substances do not elicit toxicity by themselves, they may contribute to the toxicity of mixtures. Thus, provided that water solubility is considered in the toxic unit approach, mixture toxicity should be calculated by including all PAHs present in the soil porewater. This approach presupposes that the solubility of highly lipophilic substances is not reduced in the presence of other substances, and this is supported by the results of Li and Doucette (46) that investigated the solubilities of some PCB 2434
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 11, 2002
congeners alone and in mixtures and found that, in most cases, solubility of one substance was not affected by the presence of others. In conclusion, our results show that eight highly lipophilic PAHs (log Kow > 5.6) are not toxic to springtails, whereas eight PAHs with moderate lipophilicity (log Kow ) 3.3-5.2) are. Many earlier studies have shown that toxicity of organic substances is governed by their concentration in pore-water. In the present study, a significant relationship was obtained between the log Kow and the pore-water toxicity of the eight substances that elicited toxicity, indicating a narcotic mode of toxic action. Using this relationship to estimate toxicity of the highly lipophilic PAHs, we have shown that, in most cases, absence of toxicity could be explained by the low water solubility of these substances.
Acknowledgments This study was financed by grants from the Norwegian Research Council and NorFA, and it was also supported by the Centre for Biological Processes in Contaminated Soil and Sediment (BIOPRO) and a NERI program on development of QSARs for the soil environment.
Literature Cited (1) Sverdrup, L. E.; Kelley, A. E.; Krogh, P. H.; Nielsen, T.; Jensen, J.; Scott-Fordsmand, J. J.; Stenersen J. Environ. Toxicol. Chem. 2001, 20, 1332-1338. (2) Meyer, H. Arch. Exp. Pathol. Pharmakol. 1899, 42, 109-115. (3) Verhaar, H. J. M.; Van Leeuwen, C. J.; Hermens, J. L. M. Overview of structure-activity relationships for environmental endpints: QSARs for ecotoxicity; Final report; Contract EV5V-CT92-0211; European commission, Luxembourg: Luxembourg, Belgium, 1995. (4) van Gestel, C. A. M.; Ma, W. C. Ecotoxicol. Environ. Saf. 1988, 15, 289-297. (5) van Gestel, C. A. M.; Ma, W. C. Chemosphere 1990, 21, 10231033. (6) van Gestel, C. A. M.; Ma, W. C.; Smit, C. E. Sci. Total Environ. 1991, 109/110, 589-604. (7) Ronday, R.; van Kammen-Polman, A. M. M.; Dekker: A.; Houx, N. W. H.; Leistra, M. Environ. Toxicol. Chem. 1997, 16, 601607. (8) Jager, T.; Sa´nchez, F. A. A.; Muijs, B.; van der Velde, E. G.; Posthuma, L. Environ. Toxicol. Chem. 2000, 19, 953-961. (9) Ma, W. C.; van Kleunen, A.; Immerzeel, J.; de Maagd, P. G. J. Environ. Toxicol. Chem. 1998, 17, 1730-1737. (10) Belfroid, A.; Sikkenk, M.; Seinen, W.; van Gestel, K.; Hermens, J. Environ. Toxicol. Chem. 1994, 13, 93-99. (11) Ko¨nemann, H. Toxicology 1981, 19, 209-221. (12) Mackay, D.; Shiu, W. Y.; Ma, K. C. Illustrated handbook of physical-chemical properties and environmental fate for organic chemicals. Polynuclear aromatic hydrocarbons, polychlorinated dioxins and dibenzofurans; Lewis Publishers: Chelsea, MI, 1992. (13) Wiles, J. A.; Krogh, P. H. In Handbook of soil invertebrate toxicity tests; Løkke, H., van Gestel, C. A. M., Eds.; John Wiley and sons Ltd.: Baffins, West Sussex, U.K., 1998. (14) MacFadyen, A. J. Anim. Ecol. 1961, 30, 171-184. (15) Petersen, H. Nat. Jutl. 1978, 20, 95-121. (16) Krogh, P. H.; Johansen, K.; Holmstrup, M. Appl. Soil Ecol. 1998, 7, 201-205. (17) SAS/INSIGHT User’s Guide, version 8; SAS Institute Inc.: Cary, NC, 1999. (18) SAS/STAT User’s Guide, version 8; SAS Institute Inc.: Cary, NC, 1999. (19) Kalf, D. F.; Crommentuijn T.; van de Plassche, E. J. Ecotox. Environ. Saf. 1997, 36, 89-97. (20) Krauss, M.; Wilcke, W.; Zech, W. Environ. Sci. Technol. 2000, 34, 4335-4340. (21) Baker, J. R.; Mihelcic, J. R.; Shea, E. Chemosphere 2000, 41, 813817. (22) Mackay, D.; Shiu, W. Y. J. Chem. Eng. Data 1977, 22, 399-402. (23) Walters, R. W.; Luthy, R. G. Environ. Sci. Technol. 1984, 18, 395-403. (24) Wise, S. A.; Bonnett, W. J.; Guenther, F. R.; May, W. E. J. Chromatogr. Sci. 1981, 19, 457-465.
(25) Pearlman, R. S.; Yalkowsky, S. H.; Banerjee, S. J. Phys. Chem. Ref. Data 1984, 13, 555-562. (26) Klevens, H. B. J. Phys. Colloid Chem. 1950, 54, 283-298. (27) Leo, A. J. Software CLOGP-3.42, Medchem; Medicinal Chemistry Project, Pomona College: Claremont, CA, 1986. (28) Yalkowsky, S. H.; Valvani, S. C. J. Chem. Eng. Data 1979, 24, 127-129. (29) Bayona, J. M.; Fernandez, P.; Porte, C.; Tolosa, I.; Valls, M.; Albaiges, J. Chemosphere 1999, 23, 313-326. (30) PHYSPROP database; Syracuse Research Corporation Environmental Science Center: North Syracuse, NY, 2000. (31) Freitag, D.; Ballhorn, L.; Geyer, H.; Korte, F. Chemosphere 1985, 14, 1589-1616. (32) Szabo, G.; Prosser, S. L.; Bulman, R. A. Chemosphere 1990, 21, 729-739. (33) Szabo, G.; Prosser, S. L.; Bulman, R. A. Chemosphere 1990, 21, 777-788. (34) Bowmer, C. T.; Roza, P.; Henzen, L.; Degeling, C. The development of chronic toxicological tests for PAH contaminated soils using the earthworm Eisenia fetida and the springtail Folsomia candida; TNO report IMW-R 92/387, The Netherlands, 1992. (35) Van Brummelen, T. C.; van Gestel, C. A. M.; Verweij, R. A. Environ. Toxicol. Chem. 1996, 15, 1199-1210. (36) Sverdrup, L. E.; Jensen, J.; Krogh, P. H.; Kelley, A.; Stenersen, J. Environ. Toxicol. Chem. 2002, 21, 109-114. (37) Sverdrup, L. E.; Nielsen, T.; Krogh, P. H. Relative sensitivity of three terrestrial invertebrate species to polycyclic aromatic compounds. Environ. Toxicol. Chem., in press.
(38) Sverdrup, L. E.; Ekelund, F.; Nielsen, T.; Krogh, P. H.; Johnsen, K. Soil microbial toxicity of eight polycyclic aromatic compounds: Effects on nitrification, the genetic diversity of bacteria, and the total number of protozoans. Environ. Toxicol. Chem., in press. (39) Sverdrup, L. E.; Nielsen, T.; Krogh, P. H. Toxicity of eight polycyclic aromatic compounds to red clover (Trifolium pratense), ryegrass (Lolium perenne), and mustard (Sinapsis alba). Environ. Toxicol. Chem., submitted for publication. (40) Forbes, T. L.; Forbes, V. E.; Giessing, A.; Hansen, R.; Kure L. K.. Environ. Toxicol. Chem. 1998, 17, 2453-2462. (41) Swartz, R. C.; Schults D. W.; Ozretich, R. J.; Lamberson, J. O.; Cole, F. A.; DeWitt, T. H.; Redmond, M. S.; Ferraro, S. P. Environ. Toxicol. Chem. 1995, 14, 1977-1987. (42) Ozretich, R. J.; Ferraro, S. P.; Lamberson, J. O.; Cole, F. A. Environ. Toxicol. Chem. 2000, 19, 1977-1987. (43) Alexander, M. Environ. Sci. Technol. 2000, 34, 4259-4265. (44) Kelsey, J. W.; Kottler, B. D.; Alexander, M. Environ. Sci. Technol. 1997, 31, 214-217. (45) Tang, J. X.; Alexander, M. Environ. Toxicol. Chem. 1999, 18, 2711-2714. (46) Li, A.; Doucette, W. J. Environ. Toxicol. Chem. 1993, 12, 20312035.
Received for review June 29, 2001. Revised manuscript received February 20, 2002. Accepted March 11, 2002. ES010180S
VOL. 36, NO. 11, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2435
