Environ. Sci. Technol. 2008, 42, 1332–1336
Solid-Phase Microextraction (SPME) as a Tool to Predict the Bioavailability and Toxicity of Pyrene to the Springtail, Folsomia candida, under Various Soil Conditions BJARNE STYRISHAVE,† MADS MORTENSEN,† PAUL HENNING KROGH,† O L E A N D E R S E N , ‡ A N D J O H N J E N S E N * ,† Department of Terrestrial Ecology, National Environmental Research Institute, University of Aarhus,Vejlsøvej 25, P.O. Box 314, 8600 Silkeborg, Denmark, and Department of Science, Systems and Models, University of Roskilde, P.O. Box 260, 4000 Roskilde, Denmark
Received August 22, 2007. Revised manuscript received November 13, 2007. Accepted November 14, 2007.
The porewater concentrations of pyrene were estimated by a negligible depletive solid-phase microextraction (SPME) method. The effects of organic matter (OM) and soil aging on the bioavailability of pyrene in soil were investigated by generation of reproductive effect concentrations (EC50) for the euedaphic springtail, Folsomia candida, under various test conditions. The soil used was a sandy loam soil with natural OM content of 2.6% (Askov soil). It was enriched with increasing organic matter concentrations of 5%, 10%, and 20% and was aged for 0, 56, and 112 days. The EC50 values of the springtails increased with increasing OM and aging of the soil. The increase of the OM content in the soil reduced the extractability of pyrene by SPME, as well as the toxicity of pyrene. An aging effect was demonstrated in Askov soil, EC50 values increased with increased contact time. The amounts of pyrene extracted by SPME were significantly affected by the soil concentration. Porewater concentrations determined by SPME decreased with increasing OM and aging. However, the pyrene EC50 porewater concentration remained largely constant at approximately 23 µg L-1. The results demonstrated that the SPME method is a useful indicator for bioavailability to soil microarthropods.
soil organic matter (OM). Consequently, the availability of a compound to soil organisms will depend on soil structure and mineral composition, OM content, and the physiochemical properties of the soil. As soil-pollutant contact time increases, compounds can move from the accessible soil compartments to less accessible or inaccessible compartments, which reduce extractability and bioavailability (3). Soil-compound interactions have been shown to be influenced by OM (4, 5), nanopore size, and soil structure (6). Bioavailability is governed by dynamic processes comprising several distinctive phases (7). The first processes are physicochemically driven (chemical availability) like sorption, desorption, and diffusion. These processes are controlled by substance- and soil-specific parameters such as hydrophobicity, aqueous solubility, pKa, cation exchange capacity (CEC), pH, and clay and organic matter content. The second are physiological driven uptake processes (biological availability) controlled by species-specific parameters like anatomy, surface-volume relationship, feeding strategy, and preferences in habitats. Third, there are internal allocation processes (toxicological availability) controlled by organismspecific parameters like metabolism, detoxification, storage capacity, excretion, and energy resources. Accessibility and availability are hence not necessarily identical terms, and they depend on the species in question. Arthur and Pawlinszyn (8) introduced the solid-phase microextraction (SPME) system. This device contains a small segment of fused silica fiber with a thin polymer coating for sampling of analytes. SPME has been used to estimate contaminants in sediment porewater (9, 10). This method relies on the assumption that only a small fraction of the contaminant is extracted from the porewater, and therefore, the extraction itself does not influence the equilibrium between the bound and the free form of the chemical. Exposure to contaminants through porewater is a major route of exposure for soil-dwelling organisms, and the SPME approach may therefore be a useful approach to estimate bioavailability. The present study investigates the use of SPME for the extraction of freshly spiked and aged pyrene from soil to estimate the effect of porewater concentrations and bioavailability. Furthermore the study investigates the effects of organic matter content and estimated bioavailability of pyrene on reproduction of the euedaphic springtail Folsomia candida. The experiments have been conducted over a wide range of OM content covering a span from normal agricultural soils to the high content of organic matter found in forests and other habitats throughout Europe. The applied methods might be useful for predicting the amount of pyrene available for uptake by soil-dwelling organisms and thereby improve the assessment of the actual risk of toxic compounds in soil.
Introduction As a result of fossil fuel combustion, oil spills, industrial processes, and other anthropogenic activities, polycyclic aromatic hydrocarbons (PAHs) are ubiquitous in the environment. Previous studies have shown that PAHs can be hazardous to soil-dwelling invertebrates (1). In soil, exposure of compounds to living organisms mainly occur through porewater, and toxicity therefore depends on the solubility and bioavailability of the compound in question (2). A substance can be more or less sorbed to mineral surfaces detained in micropores within the mineral fraction or in the * Corresponding author e-mail:
[email protected]; phone: +45 89201565. † University of Aarhus. ‡ University of Roskilde. 1332
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 4, 2008
Materials and Methods Experimental Soil. The Askov soil, used in the experiments, is a Danish agricultural sandy loam soil. The particle-size distribution is coarse sand (200–2000 µm), 39%; fine sand (20–200 µm), 36%; silt (2–20 µm), 12%; clay ( 0.98) from pyrene standards in methanol were used to perform quantification by measurement of peak area. Calculation of Porewater Concentrations. The total soil concentration of a chemical, Ct, is composed of a fraction in the porewater, Caq, and a fraction sorbed to soil solids (s), Cs, Ct ) (Caq · w + Cs · s)/s ) Caq · w/s + Cs ) Caq · fw + Cs
(1)
where w is water mass, s is soil particle mass, and fw is the fraction of water relative to soil. Kd, the soil–water distribution coefficient, is defined as Cs/Caq ) Kd
(2)
From this, it follows that Ct ) fw · Caq + Kd · Caq ) Caq(fw + Kd)
(3)
Caq ) Ct/(fw + Kd)
(4)
The fraction of organic carbon, fOC, is assumed to be the main sorption phase so Kd ≈ KOC · fOC
(5)
The agricultural Askov soil contained 2.6% OM corresponding to the following organic carbon content (OC), where 0.58 it the reciprocal of the van Bemmelen factor 1.726 (17) OC · 2.6% OM ) 1.5% OM
%OC ) 0.58 ·
(6)
For pyrene, log KOC ) 4.82 (1), and according to the approximate formula Kd ) KOC · fOC,
so Kd ) 104.82 · 0.015 ) 991
(7)
The water content of the original Askov soil was 16% (other OMs were 20%, 30%, and 50% water) Caq )
Ct 4.82
10
(8)
× 0.015 + 0.16
VOL. 42, NO. 4, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1333
FIGURE 1. Estimation of pyrene porewater concentrations (µg L-1) using SPME extraction after 0, 56, and 112 days of aging with varying degrees of organic matter (OM). The calculated porewater concentrations are included assuming no effects of aging in the soil. Dashed horizontal lines indicate the maximum solubility of pyrene in water (131 µg L-1). Error bars show the standard deviation (n ) 3). Determination of Porewater Concentrations with SPME. On the basis of the results from the SPME extraction, the porewater concentration could be calculated as Caq )
Cf(∞) Kf
(9)
Caq is the porewater concentration (mg L-1), Cf(∞) is the PDMS concentration on the fiber coating, and the PDMS-water partition coefficient, Kf, is 104.32 (15). Data Analysis. Estimations of the concentrations that caused 50% reduction in reproductive output (EC50 values) were related to the test concentrations and calculated in SAS/STAT, version 9.1, by the NLIN-procedure (18). ANOVA was used for a comparison of the SPME extractions.
Results Soil Concentrations over Time. The total pyrene concentration in the soil decreased with aging because of mineralization. The pyrene loss ranged from 0% to 22%. The greatest loss of pyrene during aging occurred in the soils with 10% OM, where the highest test concentration decreased from 1135 mg kg-1 (dw) at day 0 to 884 mg kg-1 (dw) at day 112. SPME Estimated Porewater Concentrations. Figure 1 shows the calculated porewater concentrations using the equations described previously and the estimated porewater concentrations using SPME extraction with varying degrees of OM in relation to aging. The figure also shows a strong correlation between the porewater concentrations estimated by SPME and the theoretically calculated porewater concentrations (P < 0.01), using the model described above. In all cases, porewater concentrations increase with increasing pyrene soil content. Within each time period, extraction with SPME fibers shows an unambiguous relationship with the soil content of organic matter (ANOVA test P < 0.01). Soils with the lowest content of OM have the highest porewater concentrations. In some cases, however, a plateau is reached at approximately 131 µg L-1, which is the water solubility of pyrene (16). Overall, the results show a decrease in the pyrene 1334
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 4, 2008
porewater concentrations with increasing organic matter content (P < 0.001). For soil with OM content of 2.6% porewater concentrations increase linearly with pyrene soil concentrations, porewater concentrations for 56 days of aging being slightly higher than 0 days and 112 days of aging (Figure 1). In the soil with 5% OM, porewater concentrations reached max solubility of approximately 131 µg L-1 at soil concentrations of 250 mg kg-1 and above. In samples with 10% OM, maximum pyrene porewater concentrations was reached at 250 mg kg-1 and above. Soils with 20% OM never reached the upper limit for pyrene solubility not even for soil aged 0 days. In these soils, the negative effect of aging on porewater concentrations is evident. In soil containing 20% OM, significantly (P < 0.05) higher measured SPME pyrene porewater concentrations was found in soil aged 0 days than in soil aged for 56 and 112 days. With this exception, no significant difference was observed between the measured SPME porewater concentrations during 0 and 56 days of aging. In total, aging exerted a negative effect on porewater concentrations (P < 0.001). With the exception of soil with 2.6% OM, the SPME measured porewater concentrations is lower in soil aged 112 days than in soil aged 0 and 56 days. Ecotoxicity Studies. The pyrene concentration causing a 50% reduction in F. candida reproduction (EC50) is shown in Figure 2. A significant increase in EC50 values with increasing content of OM was observed for all three aging periods (P < 0.05). Overall, a positive correlation between springtail EC50 and aging was also observed (P < 0.05). For the natural soil (2.6% OM), however, this correlation was not observed. The EC50 in natural Askov soil without aging (time ) 0 days) was 38.8 mg pyrene kg-1 (95% CL 31–47). Soils with 5%, 10%, and 20% OM content had an EC50 of 61.6 mg pyrene kg-1 (95% CL 49–74), 86.4 mg pyrene kg-1 (95% CL 77–96), and 160 mg pyrene kg-1 (95% CL 140–179), respectively. At a contact time of 56 days the EC50 values were significantly different (P < 0.05) except between the soil with an OM content of 2.6% and the soils with 5 and 10% OM. A factor
FIGURE 2. EC50 values for F. candida reproduction in Askov soil with varying degrees of organic matter during 0, 56, and 112 days of ageing. Error bars indicate standard errors.
FIGURE 3. Porewater EC50 concentrations (µg L-1) for pyrene in Askov soil estimated with SPME as a function of pyrene soil EC50 (mg kg-1, dw) for F. candida reproduction for all aging periods. A regression line has been included to show the relationship. Error bars indicate 95% confidence limits. of approximately three was observed between EC50 values obtained in the natural Askov soil (61 mg kg-1) and the soil with 20% OM (175 mg kg-1). The EC50 value for the natural soil (2.6% OM) was 61 mg kg-1, and the EC50 value for soil with a content of 10% OM was 106 mg kg-1, corresponding to a factor of 1.74. For data collected after 112 days, the increases in EC50 values from the Askov soil were factors of 1.6, 2.2 and 3.2 for the 5, 10 and 20% soils, respectively. Estimating EC50 Porewater Concentrations for Pyrene to Springtail Reproduction. Figure 3 shows the porewater EC50 values that can be calculated from the pyrene porewater concentrations in Figure 1 and the soil EC50 values for springtails in Figure 2. The figure demonstrates that the estimated porewater EC50 value is largely independent of soil EC50 values. OM and aging affect the pyrene porewater concentrations but not the EC50 value. Consequently, the porewater EC50 for springtail reproduction remains largely constant around 23 ( 9 µg L-1, regardless of soil concentrations, OM, and aging.
Discussion The increase OM in the soil results in lower porewater concentrations as measured by the SPME device. As the amount of OM in soil increases, there is a corresponding increase in the ability of the soil to absorb organic contaminants. This increased sorption capacity yields a reduced porewater concentration as measured by the SPME. These
results are supported by other studies. In 16 investigated soils, Chung and Alexander (19) found organic carbon as the single most important factor in the sequestration of phenanthrene, over a 200-day period. Martikainen and Krogh (20) found OM to be positively correlated to effect concentrations of dimethoate using springtails (Folsomia fimetaria) exposed in five soils with different OM content. Nam et al. (5) showed an appreciable effect of OM on sequestration in soils containing more than 2% organic carbon. When OM in soil increases, the particle surface and the number of nanopores to which pyrene can attach by adsorption and absorption increases (6). In general, two concepts have been proposed to describe the sequestration of hydrophobic organic contaminants in soil: slow diffusion into organic matter and retarded pore diffusion caused by increased sorption into soil micropores (21–23). The present results are in accordance with these concepts because they show that increased OM decreased the porewater concentration of pyrene and thereby decreased pyrene bioavailability to the springtails. A higher pyrene soil concentration is therefore needed to obtain a critical porewater concentration, leading to increased values when expressed by total soil concentrations. In contrary, EC50 expressed by porewater concentrations are largely constant at about 23 µg L-1 and is as such not affected by OM and aging. Recent work shows that not only the quantity of OM but also the quality, or more precisely specific constituents of OM, significantly contribute to strong sorption. These constituents, referred to as carbonaceous carbon phases (i.e., black carbon, coal, kerogen, or weathered oil), can be responsible for 90–99% of total sorption of organic compounds in sediment or soil (24, 25). Xing and Pignatello (26) developed a theory that conceptualizes soil organic matter as an amalgam of rubbery and glassy phases. Following this argument, absorption or partitioning processes are considered to be independent of the concentration because compounds do not compete for a limited number of sorption sites. In this study, however, the increased pyrene concentration in Askov soil significantly increased porewater concentrations until the solubility limit for pyrene in water was reached. This could be caused by pyrene competing for adsorption sites, resulting in a greater portion of the pyrene becoming available. From an ecological point of view, this finding is highly relevant because it demonstrates that the availability and hence the toxicity of pyrene is relatively higher at elevated pyrene concentrations. When soil OM content increases, the surface available for pyrene adsorption and the absorption by OM particles increases. The result appears to be a decreased effect of the pyrene concentration on the SPME extractability and the porewater concentration, that is, saturation of porewater. Consequently, the effect of the pyrene concentration and the resulting toxicity is diminished as the content of OM rises. In the Askov soil, there was a significant reduction in extraction of pyrene by SPME after 112 days of aging. It is assumed that PAH molecules in general move slowly into sites within the soil matrix that are not readily accessed (3). Organic contaminants generally exhibit two stages in the soil. The first stage occurs within minutes to hours after contact with soil where a portion of the contaminant can be quickly sorbed. The second stage where the remaining fraction is sorbed occurs on a time scale of weeks to months (21, 26, 27). Slow diffusion of the contaminant into small pores in the soil results in equally slow diffusion out of the pores and thereby they become inaccessible at least for some time. Both the mineral fraction and the organic matter contain these less-accessible pores. During soil aging, the sorption processes mentioned above become more pronounced (23). VOL. 42, NO. 4, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1335
The study demonstrates a strong correlation between the porewater pyrene concentrations estimated by SPME and the theoretically calculated porewater concentrations using the models described in eqs 1-8 above, demonstrating that both methods are useful tools in predicting porewater concentrations for pyrene and potentially other PAHs in soil. The Kf of 104.32 (15) and the KOC of 104.82 (1) used in the model to calculate porewater concentrations seem therefore valid. In conclusion, this study has demonstrated that organic matter and aging are important factors for regulation of the bioavailable fraction and toxicity of organic pollutants like PAH. It has been shown that the toxicity is proportional to porewater concentrations and is not governed by the total concentration found in the soil. The strong correlation between estimated porewater concentrations and toxicity, furthermore, confirms the conclusions made by others, for example, ref 28, that the uptake of most organic pollutants via ingested soil material seems limited for soil invertebrates. Furthermore, it was shown that SPME is a useful tool for estimation of the porewater concentrations and the effect concentrations for springtails in Askov soil. It is therefore recommended to include SPME, or other similar devises, in the risk assessment procedure of contaminated soils. How to incorporate SPME in practical terms needs further consideration. However, a few documents, for example, refs 29 and 30, include pragmatic suggestions for this.
(11) (12) (13) (14) (15)
(16)
(17)
(18) (19) (20)
Acknowledgments The Institute for Risk Assessment Sciences, Utrecht University, The Netherlands, kindly provided the PDMS fibers. Thomas ter Laak from the same institute provided fruitful comments to an early draft of the manuscript. The contribution by P.H. Krogh was supported by the NERI internal research initiative Café/NOMIRACLE.
Literature Cited (1) Sverdrup, L. E.; Nielsen, T.; Krogh, P. H. Soil ecotoxicity of polycyclic aromatic hydrocarbons in relation to soil sorption, lipophilicity, and water solubility. Environ. Sci. Technol. 2002, 36, 2429–2435. (2) Semple, K. T.; Doick, K. J.; Jones, K. C.; Burauel, P.; Craven, A.; Harms, H. Defining bioavailability and bioaccessibility of contaminated soil and sediment is complicated. Environ. Sci. Technol. 2004, 15, 228A–231A. (3) Alexander, M. Ageing, bioavailability, and overestimation of risk from environmental pollutants. Environ. Sci. Technol. 2000, 34, 4259–4265. (4) Chefetz, B.; Deshmukh, A. P.; Hatcher, P. G.; Guthrie, E. A. Pyrene sorption by natural organic matter. Environ. Sci. Technol. 2000, 34, 2925–2930. (5) Nam, K.; Chung, N.; Alexander, M. Relationship between organic matter content of soil and the sequestration of phenanthrene. Environ. Sci. Technol. 1998, 32, 3785–3788. (6) Nam, K.; Alexander, M. Role of nanoporosity and hydrophobicity in sequestration and bioavailability: Tests with model soils. Environ. Sci. Technol. 1998, 32, 71–74. (7) Lanno, R. P., Ed. Contaminated Soils: From Soil-Chemical Interactions to Ecosystem Management; SETAC Press, Pensacola, FL, 2003. (8) Arthur, C. L.; Pawliszyn, J. Solid-phase microextraction with thermal desorption using fused silica optical fibers. Anal. Chem. 1990, 62 (19), 2145–2148. (9) Mayer, P.; Vaes, W. H. J.; Wijnker, F.; Legierse, K. C. H. M.; Kraaij, R. H.; Tolls, J.; Hermens, J. L. M. Sensing dissolved sediment porewater concentrations of persistent and bioaccumulative pollutants using disposable solid-phase microextraction fibers. Environ. Sci. Technol. 2000, 34, 5177–5183. (10) King, A. J.; Readman, J. W.; Zhou, J. L. Determination of polycyclic aromatic hydrocarbons in water by solid-phase microextraction-
1336
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 4, 2008
(21) (22)
(23)
(24)
(25) (26) (27)
(28)
(29)
(30)
gas chromatography-mass spectrometry. Anal. Chim. Acta 2004, 523, 259–267. Brinch, U. C.; Ekelund, F.; Jacobsen, C. S. Method for spiking soil samples with organic compounds. Appl. Environ. Microbiol. 2002, 68, 1808–1816. Macfadyen, A. Improved funnel-type extractors for soil arthropods. J. Anim. Ecol. 1961, 30, 171–184. Petersen, H. Some properties of two high gradient extractors for soil microarthropods. Natura Jutlandica. 1978, 20, 95–121. Krogh, P. H.; Johansen, K.; Holmstrup, M. Automatic counting of collembolans for laboratory experiments. Appl. Soil Ecol. 1998, 7, 201–205. Ter Laak, T. L.; Barendregt, A.; Hermens, J. L. M. Freely dissolved pore water concentrations and sorption coefficients of PAH in spiked, aged, and field-contaminated soils. Environ. Sci. Technol. 2006, 40, 2184–2190. Ter Laak, T. L.; Agbo, S. O.; Barendregt, A.; Hermens, J. L. M. Freely dissolved concentrations of PAHs in soil pore water: Measurements via solid-phase extraction and consequences of soil tests. Environ. Sci. Technol. 2006, 40, 1307–1313. Westman, C. J.; Hytönen, J.; Wall, A. Loss-on-ignition in the determination of pools of organic carbon in soils of forests and afforested arable fields. Commun. Soil Sci. Plant Anal. 2006, 37, 1059–1075. SAS/STAT User’s Guide, version 9.1; SAS Institute, Inc.: Cary, NC, 2004. Chung, N.; Alexander, M. Effect of soil properties on bioavailability and extractability of phenanthrene and atrazine sequestered in soil. Chemosphere 2002, 48, 109–115. Martikainen, E. A. T.; Krogh, P. H. Effects of soil organic matter content and temperature on toxicity of dimethoate to Folsomia fimetaria (Collembola: Isotomiidae). Environ. Toxicol. 1999, 18, 865–872. Pignatello, J. J.; Xing, B. Mechanisms of slow sorption of organic chemicals to natural particles. Environ. Sci. Technol. 1996, 30, 1–11. Cornelissen, G.; Vann Nort, P. C. M.; Govers, H. J. Mechanism of slow desorption of organic compounds from sediments: A study using model sorbents. Environ. Sci. Technol. 1998, 32, 3124–3131. Semple, K. T.; Morris, A. W. J.; Paton, G. I. Bioavailability of hydrophobic organic contaminants in soils: Fundamental concepts and techniques for analysis. Eur. J. Soil Sci. 2003, 54, 809–818. Cornelissen, G.; Gustafsson, Ö.; Bucheli, Th.D.; Jonker, M. T. O.; Koelmans, A. A.; van Noort, P. C. M. Critical review: Extensive sorption of HOCs to black carbon, coal and kerogen: Mechanisms and consequences for sorption, bioaccumulation and biodegradation. Environ. Sci. Technol. 2005, 39, 6881–6895. Koelmans, A. A.; Jonker, M. T. O.; Cornelissen, G.; Bucheli; Th, D.; van Noort, P. C. M.; Gustafsson, 951 > O. Black Carbon: The Reverse of its Dark Side. Chemosphere 2006, 63, 365–377. Xing, B.; Pignatello, J. J. Dual-mode sorption of low-polarity compounds in glassy polyvinyl chloride and soil organic matter. Environ. Sci. Technol. 1997, 31, 792–799. Reid, B. J.; Jones, K. C.; Semple, K. T. Bioavailability of persistent organic pollutants in soils and sedimentsA perspective on mechanisms, consequences and assessment. Environ. Pollut. 2000, 108, 103–112. Belfroid, A. C.; Seinen, W.; Van Gestel, C. A. M.; Hermens, J. L. M.; Van Leeuwen, K. J. Modelling the accumulation of hydrophobic organic chemicals in earthwormssApplication of the equilibrium partitioning theory. Environ. Sci. Poll. Res. 1995, 2, 5–15. Jensen, J.; Mesman, M. Ecological Risk Assessment of Contaminated Land. Decision Support for Site Specific Investigations; RIVM: Bilthoven, The Netherlands, 2006; ISBN 90-6960-138-9, 136 pp. Van der Wal, L.; Jager, T.; Fleuren, R.H.L.J.; Barendregt, A.; Sinnige, T. L.; Van Gestel, C. A. M.; Ter Laak, T. L.; Durjana, M.; Struijs, J.; Hermens, J. L. M. Solid-phase dosing and sampling technique to determine partition coefficients of hydrophopic chemicals in complex matrixes. Environ. Sci. Technol. 2005, 39, 3736–3742.
ES072102W
