Archives of Environmental Contamination and Toxicology

, Volume 61, Issue 1, pp 74–82

Bioavailability of Hydrophobic Organic Contaminants in Sediment with Different Particle-Size Distributions


  • W. Tyler Mehler
    • State Key Laboratory of Organic Geochemistry, Guangzhou Institute of GeochemistryChinese Academy of Sciences
    • Fisheries and Illinois Aquaculture Center and Department of ZoologySouthern Illinois University
  • Huizhen Li
    • State Key Laboratory of Organic Geochemistry, Guangzhou Institute of GeochemistryChinese Academy of Sciences
    • Graduate SchoolChinese Academy of Sciences
  • Junxiao Pang
    • State Key Laboratory of Organic Geochemistry, Guangzhou Institute of GeochemistryChinese Academy of Sciences
    • Graduate SchoolChinese Academy of Sciences
  • Boquan Sun
    • State Key Laboratory of Organic Geochemistry, Guangzhou Institute of GeochemistryChinese Academy of Sciences
    • Graduate SchoolChinese Academy of Sciences
  • Michael J. Lydy
    • Fisheries and Illinois Aquaculture Center and Department of ZoologySouthern Illinois University
    • State Key Laboratory of Organic Geochemistry, Guangzhou Institute of GeochemistryChinese Academy of Sciences

DOI: 10.1007/s00244-010-9609-z

Cite this article as:
Mehler, W.T., Li, H., Pang, J. et al. Arch Environ Contam Toxicol (2011) 61: 74. doi:10.1007/s00244-010-9609-z


Few studies have been conducted examining the distribution of different-sized particles in sediment and its potential impact on bioavailability of sediment-associated contaminants. In the current study, three sediments composed of different particle sizes, i.e., fine (0–180 μm), combined (0–500 μm), and coarse (180–500 μm), were used to evaluate the bioaccumulation potential and toxicokinetic rates of four hydrophobic organic contaminants (HOCs) including two polychlorinated biphenyls (PCB-101 and PCB-118), a metabolite of an organochlorine insecticide (p,p′-DDE), and a polybrominated diphenyl ether (BDE-47) to the benthic oligochaete Lumbriculus variegatus. Two chemical approaches, Tenax extraction and matrix-solid phase microextraction (SPME), were also used to measure bioavailability of the sediment-associated HOCs. The uptake and elimination rates of HOCs by L. variegatus from coarse sediment were greater than those from fine sediment, although the biota–sediment accumulation factors (BSAFs) were not significantly different among sediments with different particle sizes. The freely dissolved HOC concentrations measured by matrix-SPME were greater in coarse sediment, however, no difference was found in uptake and desorption rates for the matrix-SPME and Tenax extraction measurements. Although BSAFs in L. variegatus were the same among sediments, kinetic rates of HOCs for organisms and freely dissolved HOC concentrations were lower in fine sediment, suggesting that sediment ingestion may also play a role in organism uptake, especially for HOCs in fine sediment.

Traditionally, assessing risk caused by hydrophobic organic contaminants (HOCs) in sediment has been based on total sediment concentrations measured by exhaustive chemical extraction. However, using total extractable concentrations may overestimate the risk due to the limited bioavailability of HOCs (Jonker et al. 2007; You et al. 2008). A variety of factors, such as biological features of the organisms, chemical properties, chemical–sediment contact time, and sediment characteristics (Harkey et al. 1994; Hendriks et al. 2001; Cornelissen et al. 2005; Kukkonen et al. 2005; You et al. 2007a; Sormunen et al. 2008; Trimble et al. 2008), may affect bioavailability of HOCs in sediment.

Sediments are not homogenous but rather contain multiple compartments differing dramatically in regard to physical and chemical composition. Although the role of organic carbon (OC) and other sediment composition factors, such as black carbon and sedimentary plant-derived carbon, have been well studied with respect to bioavailability (Rochne et al. 2002; Cornelissen et al. 2005; Kukkonen et al. 2005; You et al. 2006; Yang et al. 2008), few studies have investigated the role that physical properties, such as sedimentary particle-size distribution, may play in this estimation (Cornelissen et al. 1999; Millward et al. 2001; Laak et al. 2007).

Previous studies have suggested that desorption and porewater concentrations of sediment-associated polycyclic aromatic hydrocarbons (PAHs) changed when sediment particle size changed (Cornelissen et al. 1999; Laak et al. 2007). Moreover, selective ingestion of fine sediment by benthic organisms also make particle-size distribution an important factor affecting bioavailability of HOCs in sediment (Millward et al. 2001). In addition, particle sizes equate to particle transport rates in the aquatic environment. Coarse particles are usually transported by way of bottom currents, whereas the lighter-fine particles are resuspended into the upper water column (Eisma 1993). The different transport mechanisms can result in different size fractioned sediments depending on locations in the water body. For these reasons, understanding how particle-size distribution influences HOC bioavailability is important for assessing the risks of contaminants in both laboratory and field assessments.

Non-exhaustive chemical extraction has been shown to be a more accurate method for predicting bioavailability and toxicity of sediment-associated HOCs compared with exhaustive extraction methods because it estimates only the bioavailable fractions within the system (Conder et al. 2004; You et al. 2006; Jonker et al. 2007; You et al. 2007b; Styrishave et al. 2008; van der Heijden and Jonker 2009; Oleszczuk 2009), and it has been effective in evaluating bioavailability of various classes of HOCs in sediment with different OC (You et al. 2006; Jonker et al. 2007) and black carbon contents (Cornelissen et al. 2005; Yang et al. 2008). However, relatively few studies have examined whether non-exhaustive extraction techniques can accurately predict the bioavailable fractions of HOCs when differing physical factors, such as particle-size distribution, occur.

The objectives of the current study were to evaluate bioaccumulation potential and processes of four HOCs including two polychlorinated biphenyls (2,2′,4,5,5′-pentachlorobiphenyl [CB-101] and 2,3,4,4′,5′-pentachlorobiphenyl [CB-118]), a metabolite of an organochlorine insecticide (4,4′-dichlorodiphenyldichloroethylene [DDE]), and a polybrominated diphenyl ether (3,3′,4,4′-tetrabromodiphenyl ether [BDE-47]) in three sediments with different particle-size distributions to Lumbriculus variegatus. The four HOCs are ubiquitous in the environment and have log Kow values of 6.38, 6.74, 6.51, and 6.81 for CB-101, CB-118, DDE, and BDE-47, respectively (Verschueren 1983; Hawker and Connell 1988; Braekevelt et al. 2003). In addition, two non-exhaustive chemical extraction techniques, Tenax extraction and matrix-solid phase microextraction (SPME), were used to evaluate bioavailability of the sediment-associated HOCs from the three sediments by measuring the kinetic rates and concentrations at equilibrium. The capability of using these two chemical approaches for measuring bioavailability of HOCs was also evaluated compared with previous studies.

Materials and Methods

Chemicals and Sediments

Two penta-CBs (CB-101 and CB-118) were purchased from Accustandard (New Heaven, NJ), and DDE and BDE-47 were obtained from ChemService (West Chester, PA). Decachlorobiphenyl (DCBP) and 4,4′-dibromoocta-fluorobiphenyl (DBOFB) were used as surrogates and added to the samples before extraction to verify the performance of the analytical processes and were purchased from Supelco (Bellefonte, PA), and CB-24 and CB-189 (Accustandard) were added to the cleaned extracts and used as internal standards for gas chromatography–mass spectrometry (GC–MS) quantification. Sodium azide (NaN3, analytical grade) was used to prevent microbial growth during the biomimetic extraction experiments and was purchased from Lianhe Chemical Company (Chengdu, China).

Three sediments were used, namely, fine, combined, and coarse sediments, which were composed of particles ranging in size from 0 to 180, 0 to 500, and 180 to 500 μm, respectively. The sediments were prepared from a clean soil collected from Baiyu Mountain in Guangzhou, China. No target compounds were detected in the soil. The soil was ground and passed through a series of sieves to obtain the desired size fractions and then hydrated to make sediment using moderately hard water (MHW), which was made according to United States Environmental Protection Agency protocols (USEPA 2000). The mean dry-to-wet ratio of the sediments was 61.0% ± 0.013%, and total OC contents of the three sediments were analyzed using an Elementar Vario ELIII (Hanau, Germany).

The four contaminants were spiked into sediment at environmentally realistic but high enough concentrations to be quantifiable for body residues and for successful use of the biomimetic extraction techniques. Acetone was used as carrier for sediment spiking. After spiking, sediments were thoroughly mixed for 4 h using a drill with a rotating stainless-steel blade. The spiked sediments were stored at 4°C for 45 days and re-homogenized before use.


The oligochaete worm L. variegatus were cultured in Guangzhou Institute of Geochemistry, China, according to USEPA standard protocols (United States Environmental Protection Agency 2000) and were used in the bioaccumulation tests. Bioassays were conducted using five replicates in 400-ml beakers. Each beaker contained 80 g wet sediment and approximately 200 ml overlying water, and the water was changed twice a day (150 ± 50 ml) using an automated water-delivery system. After sediment was allowed to settle overnight, 20 worms were placed into each beaker. Testing was conducted using a 16:8–h light-to-dark photoperiod and at 21°C ± 2.5°C; temperature, conductivity, pH, and dissolved oxygen (DO) were measured daily.

Bioaccumulation testing was terminated at 72, 168, 240, 336, and 672 h by sieving the L. variegatus from sediment with a 500-μm sieve. Worms were transferred into clean MHW for 6 h to allow for gut purging. The organisms were blotted dry and then weighed using a Sartorius Ag Pro 11 microbalance (Gottingen, Germany). One worm from each replicate was used for lipid analysis according to a spectrophotometric method after acid digestion (Handel 1985), whereas the remaining worms were frozen at –20°C for subsequent chemical body residue analysis by GC–MS after sonication extraction and sulfuric acid cleanup. After adding 50 ng each surrogate (DBOFB and DCBP), organisms were extracted with acetone in five cycles of 10 s each at 600 W using an ultrasonic processor (Scientz Biotechnology, Ningbo, China). The extracts were filtered, solvent-exchanged to hexane, cleaned using concentrated sulfuric acid, and then evaporated to 1 ml under a gentle stream of nitrogen. Finally, 50 ng of each internal standard (CB-24 and CB-189) was added to the extracts before GC–MS analysis.

In addition to bioaccumulation testing, sublethal toxicity was examined after the 672 h (28 days) of bioassay by recording worm mass, reproductive success, and dorsal blood vessel values (Mäenpää et al. 2009). Dorsal blood vessel rate was assessed in triplicate with five organisms analyzed in each replicate. The organism was fixed into a small slit in a Parafilm (American Can, Neenah, WI, USA) coated microscope slide, and the pulses were counted under a XSP-136 binocular microscope (Phenix Optics, Shangrao, China).

Consecutive Tenax Extraction

Consecutive Tenax extraction was performed in triplicate to measure desorption rates of HOCs from sediments, and six sampling time points were used (3, 6, 24, 72, 168, and 336 h) (You et al. 2007b). In brief, 5 g wet sediment, 5 ng NaN3, 45 ml MHW, and 0.5 g Tenax beads were added into a 50-ml screw-cap glass tube. The tubes were rotated at 20 rpm on a QB-228 tube rotator (Kylin-Bell, Haimen, China). At each pre-determined sampling time point, Tenax beads were removed from the sediment slurry, and fresh Tenax was added to resume the experiment. Tenax beads were extracted by sonication three times with 5 ml of a hexane-and-acetone (1:1, v/v) solution, and the three extracts were combined. After adding 50 ng surrogates (DBOFB and DCBP), the extracts were cleaned with 1 ml concentrated sulfuric acid and filtered through a column packed with anhydrous Na2SO4. Eluents were then concentrated to 0.5 ml under a gentle stream of nitrogen, and 50 ng CB-24 and CB-189 were added as the internal standards for GC–MS quantification.


Matrix-SPME was used to measure freely dissolved contaminants in sediment porewater, and these experiments were conducted in triplicate. In the current study, disposable SPME fibers coated with 30 μm polydimethylsiloxane were used and had a phase volume of 0.124 μl/cm (PolyMicro, Phoenix, AZ). The fibers were sampled at five time points (72, 168, 240, 336, and 672 h) to ensure that the HOCs had reached equilibrium between the sediment porewater and the fibers. Matrix-SPME experiments were performed according to previously published procedures (You et al. 2007b). The fibers were protected with a 2 × 2-cm envelope made of stainless steel mesh with 110-μm openings. Before use, the envelopes containing the fibers were sonicated in methanol for 10 min using a KH5200D water-bath sonicator (Hechuang Sonication Instrument, Kunshan, China), washed with deionized water, and dried. Each envelope of 5 cm fibers was inserted into 10 g wet sediment in 60-ml glass vials and shaken gently on a HY-4 shaker (Fuhua Instrument, Jintan, China). At each time point, three vials were taken off of the shaker and the fibers removed from the sediment. After washing with deionized water and being dried, the fibers were sonicated in 1 ml hexane for 5 min, and the extraction was repeated two additional times. Extracts were combined and concentrated to 100 μl, with 50 ng CB-24 and CB-189 added as internal standards before analysis with GC–MS.

Chemical Analysis

Total HOCs were extracted from the sediments using a CW-2000 ultrasound-assisted microwave extractor (Xintuo, Shanghai, China). Extracts were cleaned with concentrated sulfuric acid and analyzed by GC–MS. In short, 5 g freeze-dried sediment and 2 g activated copper powder were placed in 250-ml extraction flasks. After adding 50 ng surrogates, extraction was initiated using a 100 ml mixture of hexane and acetone (1:1, v/v). The extraction time and microwave power were set at 6 min and 100 W, respectively. The extracts were filtered, concentrated, and solvent-exchanged to 1 ml hexane using a rotary evaporator (IKA, Staufen, Germany). Extracts were cleaned with 1 ml concentrated sulfuric acid, dried with anhydrous Na2SO4, concentrated to 1 ml hexane, and analyzed with GC–MS after adding 50 ng internal standards.

Target analytes in the extracts were analyzed using a Shimadzu QP 2010 Plus series GC–MS (Shimadzu, Kyoto, Japan) in negative chemical ionization (NCI) mode, and chemical separation was achieved using a Rtx-5MS column (30 m × 0.25 mm, 0.25-μm film thickness). Helium was used as carrier gas at a flow rate of 1 ml/min, and methane was used as NCI reaction gas. The ion source and transfer line temperatures were set at 250°C and 260°C, respectively. Oven temperature was 108°C for 1 min, heated to 220°C at 18°C/min, heated to 250°C at 8°C/min, heated to 255°C at 1°C/min, then heated to 280°C at 5°C/min and held at 280°C for 5 min. The cleaned extract (1 μl) was injected using a pulsed splitless mode, and the injector temperature was set at 260°C. Selective ion monitoring was used, and the most abundant ion in the scan mode (m/z 35 for CBs and DDE and m/z 71 for BDE 47, respectively) was selected as the target ion for each compound. Identification of analytes was based on detection of target ion and qualifiers within the 1% retention time window, and internal standard calibration was used for chemical quantification with CB-24 and CB-189 used as internal standards.

Data Analysis

The kinetic rates and chemical body residues in L. variegatus at steady state and on the SPME fibers at equilibrium were estimated using a first-order one-compartment model (Eq. 1) with Scientist 2.01 (MicroMath Scientific, St. Louis, MO) as follows:
$$ C_{\text{t}} = C\left( { 1- {\text{e}}^{{-k{\text{e}}\cdot{\text{t}}}} } \right) $$
where Ct and C represent the HOC concentration in the worm or fiber at time t (h) and at steady state (at equilibrium), respectively, and ke is the elimination rate constant or fiber desorption rate constant (1/h) for bioaccumulation testing and matrix-SPME, respectively.
Desorption rates and rapid desorption fractions (Fr) estimated by Tenax extraction were fit to a triphasic model (Eqs. 2 and 3) using Scientist 2.01 as follows:
$$ S_{\text{t}} /S_{0} = F_{\text{r}} \left( {{\text{e}}^{{ - k{\text{r}}\cdot{\text{t}}}} } \right) \, + F_{\text{s}} \left( {{\text{e}}^{{ - k{\text{s}}\cdot{\text{t}}}} } \right) \, + F_{\text{vs}} \left( {{\text{e}}^{{ - k{\text{vs}}\cdot{\text{t}}}} } \right) $$
$$ F_{\text{r}} + F_{\text{s}} + F_{\text{vs}} = { 1} $$
where St and S0 represent the amount of sediment-sorbed chemical at time t (h) and time zero, respectively; Fr, Fs, and Fvs are the fraction of chemical in the rapidly, slowly, and very slowly desorbing fractions at time zero; and kr, ks and kvs are the corresponding desorption rate constants (1/h), respectively.
Bioaccumulation potential of HOCs from the sediments to L. variegatus was described through biota–sediment accumulation factors (BSAFs), which were calculated by dividing the lipid-normalized biota concentration (Cb) by the OC-normalized sediment concentration (Cs) as follows (Eq. 4):
$$ {\text{BSAF }} = C_{\text{b}} ({\text{lipid}} - {\text{normalized}}) \, /C_{\text{s}} ({\text{OC}} - {\text{normalized}}) $$
The BSAF equals the ratio of the uptake (ku) and elimination (ke) rates; therefore, ku (g OC/g lipid/h) was calculated as Eq. 5 as follows:
$$ k_{\text{u}} = {\text{ BSAF}} \bullet k_{\text{e}} $$
Sediment porewater concentration (Cpw) was calculated by dividing fiber concentration (Cf) by the partition coefficient between the fiber and the water (Kfw), which was calculated using the relationship between Kfw and Kow (log Kfw = log Kow − 0.91) (Mayer et al. 2000) as follows (Eq. 6):
$$ C_{\text{pw}} = {\frac{{C_{\text{f}} }}{{K_{\text{fw}} }}} $$

All statistical comparisons were conducted using analysis of variance (α = 0.05) and either Tukey’s Honestly Significant Difference or Student t test employing SAS 9.1 software (SAS, Cary, NC).

Results and Discussion


All measured water-quality parameters were within USEPA guidelines, with the exception that the temperature (21°C ± 2.5°C) used in the current study was a little lower than that stated in the USEPA guidelines (i.e., 23°C ± 1°C) (United States Environmental Protection Agency 2000). No avoidance of sediment was observed for any of the worms during testing. In addition, sublethal toxicity to L.variegatus including mass, reproduction, and dorsal blood vessel rates were assessed at the end of the 28-day exposure (Mäenpää et al. 2009). No significant differences were noted between controls and treatments for any of the sublethal endpoints (data not shown). Furthermore, there were no significant differences (F14,59 = 1.06, p = 0.41) in lipid content in the worms among controls and treatments; thus, the average lipid contents (1.82% ± 0.26%) were used. In summary, the sublethal endpoints and lipid data suggest that the body residues achieved in the worms during the bioaccumulation tests were not great enough to adversely affect the bioaccumulation results.

The bioaccumulation potential and process, expressed as BSAFs, and the toxicokinetic rates of the four HOCs from the three sediments to L.variegatus are listed in Table 1. The BSAF values were calculated by dividing the lipid normalized chemical body residues at steady state by the OC-normalized sediment concentrations. Initial sediment concentrations were 198 ± 5.3, 186 ± 5.8, 245 ± 5.0, and 219 ± 7.4 ng/g dry weight for CB-101, CB-118, DDE, and BDE-47, respectively, with no significant differences being noted in concentrations among the different size fractioned sediments. No degradation of HOCs was observed throughout the testing because concentrations at the end of testing were within 80–120% of initial concentrations. Therefore, the average values of the measured HOC concentrations at the beginning and the end of testing were used as sediment concentrations. Fine sediment contained significantly greater amounts of OC (1.76% ± 0.03%) than did both combined and coarse sediments (1.54% ± 0.04% and 1.53% ± 0.10%, respectively), and no significant difference was noted between the latter two sediments.
Table 1

BSAFs, ke, and ku of target contaminants to L. variegatus exposed to fine (0–180 μm), combined (0–500 μm), and coarse (180–500 μm) sedimentsa


BSAF (g OC/g lipid)b

ke (1/h)

ku (g OC/g lipid/h)











4.79 ± 0.45a

5.01 ± 0.46a

4.86 ± 0.41a

0.007 ± 0.002a

0.009 ± 0.002ab

0.012 ± 0.003b

0.033 ± 0.008a

0.045 ± 0.011ab

0.056 ± 0.017b


2.95 ± 0.74a

3.02 ± 0.40a

3.38 ± 0.43a

0.006 ± 0.002a

0.009 ± 0.003ab

0.012 ± 0.005b

0.019 ± 0.007a

0.027 ± 0.011ab

0.041 ± 0.017b


10.36 ± 3.43a

8.17 ± 1.53a

6.99 ± 0.96a

0.002 ± 0.001a

0.004 ± 0.001b

0.006 ± 0.002b

0.024 ± 0.013a

0.034 ± 0.013ab

0.044 ± 0.015b


4.37 ± 0.66a

4.56 ± 0.73a

4.47 ± 0.58a

0.005 ± 0.001a

0.006 ± 0.002ab

0.008 ± 0.003b

0.020 ± 0.006a

0.028 ± 0.009ab

0.037 ± 0.013b

aData are presented as means ± SDs

bBSAF = lipid normalized biota concentration/organic carbon normalized sediment concentration. Different superscript letters indicate significant differences among treatments

The performance of the analytical procedures for extracting and cleaning the HOCs from the worms was assessed, and recoveries were 85.1% ± 17.5%, 81.1% ± 10.9%, 87.8% ± 14.4%, and 94.3% ± 26.6% for CB-101, CB-118, DDE, and BDE-47, respectively. Body residues at steady state, which were estimated by fitting HOC data at 72, 168, 240, and 336 h to the first-order one-compartment model, were used to calculate BSAFs. It should be noted, however, that the 672-h time point in the L. variegatus bioaccumulation testing was not used because chemical body residues decreased dramatically at 28 days compared with the 14-day results. Similar findings have been reported with oligochaete accumulation being greatly decreased when exposure times increased from 14 to 28 days (You et al. 2006). It has been suggested that architometric reproduction (e.g., fragmentation), which halts feeding behavior in worms, may be the reason for the decreased bioaccumulation (Leppanen and Kukkonen 2000; You et al. 2006).

As listed in Table 1, BSAF values varied among the four compounds, with DDE having the largest BSAFs, which ranged from 6.99 to 10.4 among the three sediments. The DDE BSAFs were higher than that in a study by You et al. (2006), who reported BSAF values for DDE as high as 5.31. The DDE body residues at 14 days were used to calculate the BSAFs in You et al. (2006), whereas body residues at steady state, which were estimated by the uptake model, were used in the current study. The BSAF values at steady state for the PCBs and PBDE ranged from 2.95 to 5.01, and the lowest BSAFs were for the planar CB-118 (Table 1). When BSAF equals one, sediment OC and organism lipid have the same affinity for the HOCs. With BSAF values being greater than one, the target contaminants showed a higher affinity to organism lipid compared with sediment OC, indicating greater bioaccumulation potential. Although differences in BSAFs were evident among compounds, no significant differences were noted for a single compound in sediments with different particle-size distributions.

The toxicokinetic rates of the HOCs from the three sediments in the bioaccumulation testing were estimated as well. Compared with BSAF values, significant differences existed for the uptake (ku) and elimination (ke) rates of the same compound from sediments with different particle-size distributions (Table 1). In general, chemicals in coarse sediment had greater ku and ke values, whereas the toxicokinetic processes of the worms were slower for chemicals in the fine sediment (Table 1). The L. variegatusku and ke values in the combined sediment were not significantly different from those in the fine or coarse sediment (with the exception of ke for DDE in the combined sediment, which was significantly greater than that in the fine sediment).

Non-exhaustive Chemical Techniques

In addition to bioaccumulation testing with L. variegatus, Tenax extraction and matrix-SPME measurements were conducted to measure desorption of HOCs from the sediment (Table 2) and the freely dissolved chemical concentrations in sediment porewater (Cpw) (Table 3), respectively. Tenax-extraction measurements were fit to a triphasic model, and the coefficients of determination (COD) values were all good, with values >0.97 for the four compounds in all sediments. Uptake of the target compounds by the SPME fibers were fit to a first-order one-compartment model, and the COD values ranged from 0.36 to 0.83. The lower COD values and greater SDs associated with the matrix-SPME measurements may partially be explained by the difficulty of measuring low concentrations of HOCs extracted by the SPME fibers. As listed in Tables 2 and 3, no significant differences were observed in desorption or uptake rates of sediment-associated HOCs measured by either chemical technique among the three sediments with different particle-size distributions. However, Cpw, measured by matrix-SPME at equilibrium, was significantly different among the three sediments (Table 3). With the exception of CB-118, Cpw of the other three compounds was significantly larger in the coarse sediment than those in the fine sediment, which agreed with the kinetic rates in the L. variegatus bioaccumulation test. Similarly, Cpw of HOCs in the combined sediment was not significantly different from either the coarse or fine sediment for the four compounds tested, with the exception that the Cpw of CB-101 in the combined sediment was greater than that in fine sediment (Table 3).
Table 2

The Fr, Fs, and Fvs of the target contaminants from fine (0–180 μm), combined (0–500 μm), and coarse (180–500 μm) sediments and their corresponding desorption coefficients kr, ks, and kvs, respectivelya















0.28 ± 0.11

0.30 ± 0.019

0.24 ± 0.041

0.48 ± 0.088

0.52 ± 0.018

0.53 ± 0.036

0.24 ± 0.14

0.18 ± 0.021

0.23 ± 0.049


0.24 ± 0.076

0.28 ± 0.075

0.22 ± 0.042

0.38 ± 0.071

0.38 ± 0.086

0.42 ± 0.048

0.39 ± 0.093

0.33 ± 0.074

0.36 ± 0.046


0.27 ± 0.091

0.29 ± 0.025

0.21 ± 0.034

0.42 ± 0.065

0.43 ± 0.023

0.47 ± 0.028

0.32 ± 0.12

0.28 ± 0.028

0.32 ± 0.042


0.24 ± 0.058

0.27 ± 0.049

0.19 ± 0.027

0.36 ± 0.048

0.39 ± 0.047

0.40 ± 0.028

0.41 ± 0.031

0.34 ± 0.057

0.42 ± 0.031














CB- 101

0.017 ± 0.017

0.021 ± 0.003

0.019 ± 0.008

0.31 ± 0.091

0.38 ± 0.024

0.33 ± 0.039

0.00034 ± 0.0019

0.00058 ± 0.0004

0.00032 ± 0.0007

CB- 118

0.018 ± 0.015

0.023 ± 0.016

0.021 ± 0.011

0.27 ± 0.080

0.31 ± 0.11

0.27 ± 0.041

0.00021 ± 0.0008

0.00030 ± 0.0008

0.00024 ± 0.0005


0.016 ± 0.013

0.020 ± 0.004

0.017 ± 0.007

0.29 ± 0.069

0.37 ± 0.028

0.31 ± 0.032

–0.00004 ± 0.0011

0.00008 ± 0.0003

–0.00001 ± 0.0004


0.017 ± 0.011

0.021 ± 0.009

0.020 ± 0.007

0.28 ± 0.055

0.35 ± 0.071

0.29 ± 0.029

0.00040 ± 0.0006

0.00053 ± 0.0005

0.00038 ± 0.0003

aData are presented as means ± SDs. No significant differences were noted for Tenax measurement among the three sediments

Table 3

Cpw of the target contaminants in fine (0–180 μm), combined (0–500 μm), and coarse (180–500 μm) sediments measured by matrix-SPMEa


Cpw (ng/l)

ke (1/h)








7.32 ± 1.70a

12.0 ± 2.70b

12.1 ± 1.78b

0.0070 ± 0.0049a

0.0027 ± 0.0011a

0.0027 ± 0.0011a


2.43 ± 0.47a

5.45 ± 1.89a

3.45 ± 0.50a

0.0057 ± 0.0030a

0.0016 ± 0.0008a

0.0016 ± 0.0008a


4.95 ± 1.18a

6.19 ± 0.77ab

7.92 ± 1.00b

0.0079 ± 0.0061a

0.0057 ± 0.0019a

0.0057 ± 0.0019a


1.41 ± 0.21a

1.81 ± 0.27ab

1.99 ± 0.21b

0.0083 ± 0.0042a

0.0061 ± 0.0025a

0.0061 ± 0.0025a

aThe ke of the contaminants with respect to the SPME fibers are also presented. Data are presented as means ± SDs. Different superscript letters indicate significant differences among treatments

In summary, uptake and elimination rates for L. variegatus during the bioaccumulation tests were faster in the coarse sediment and were the slowest in the fine sediment. A similar trend was noted in porewater concentrations of the HOCs in the three sediments. This result makes sense because uptake and, potentially, elimination will be regulated by the amount of compound freely dissolved in the porewater. However, no significant difference was observed in the BSAF values for a single compound from different sediments. This result suggests that uptake of HOCs from porewater was not the sole uptake route and that sediment ingestion also played an important role in bioaccumulation of HOCs, especially from fine sediment. In addition, BSAF measurements are made at steady state and include both uptake and elimination components, which could potentially cancel each other out.

Role of Particle-Size Distribution in Sediment Risk Assessment

The heterogeneity of sediments can greatly influence bioavailability of sediment-associated contaminants and this fact has been well documented; however, the role of particle-size distribution is still rarely addressed in current risk assessments. Fine sediment particles contain higher OC content and greater surface area than larger particles; thus, they potentially have stronger sorption to HOCs, which was indicated by the lower Cpw found in the current study (Table 3). In contrast, many benthic organisms selectively feed on fine particles, and this may lead to higher contaminant loads for organisms exposed to fine sediment (Harkey et al. 1994). Furthermore, in sandy sediment, which is composed of larger-sized particles, lower bioavailability has been noted for HOCs (Amweg et al. 2006; Trimble et al. 2008) coupled with slower desorption of the HOCs from sand (Cornelissen et al. 1999; Trimble et al. 2008). Thus, evaluating sediment particle-size distribution, along with organisms’ feeding behavior, could greatly improve accuracy in assessing risks due to sediment-associated contaminants.

Although in the current study no significant difference was found in BSAF values for the target compounds among the three sediments, the toxicokinetic rates still varied (Table 1). Slower uptake and elimination of HOCs by L. variegatus were observed in fine sediment, which had a higher OC content. Although the differences in the rates among treatments were subtle, they could affect risks of HOCs to aquatic organisms. In the current study, no toxic effects were noted (lethal or sublethal) due to the presence of the contaminants in sediment regardless of particle-size distribution. However, in sediment with higher contamination loadings than used in the current study or in sediment contaminated by more toxic compounds, the differences in uptake rates induced by different sediment particle-size distribution may impact the survival of the exposed organisms.

Because the toxicokinetic rates for L. variegatus were significantly different among sediments with different particle sizes, Tenax extraction and matrix-SPME were used to measure desorption rates of HOCs from sediment and uptake rates of HOCs to SPME fibers, respectively. However, no significant differences in the kinetics rates were detected for either chemical approach (Tables 2, 3). In fact, the pattern of increased toxicokinetic rates with increased particle sizes observed in the L.variegatus bioaccumulation testing were not observed using either chemical approach. These results suggested that other routes of exposure (i.e., sediment ingestion) may play a role in the toxicokinetic processes in the bioaccumulation testing.

Non-exhaustive Extractions as Surrogates for Measuring Bioavailability

Recent studies have examined Tenax extraction and matrix-SPME as surrogates for measuring bioavailability of various HOCs (including PCBs, PAHs, and pesticides) from sediments containing different TOC levels, and strong positive correlations have been reported between chemical and biological availability (You et al. 2006; Trimble et al. 2008; van der Heijden and Jonker 2009). The relationship between chemical measurements (Tenax-extractable sediment concentrations and SPME fiber concentrations) and biological measurements (HOCs’ body residues) from the current study in addition to data from previous studies are presented in Figs. 1 and 2.
Fig. 1

The relationship between 6-h Tenax extraction (log Cs6, ng/g OC) to L. variegatus bioaccumulation testing (log Cb, ng/g lipid) using data from previous studies (open triangle) (You et al. 2006; You et al. 2007a; Trimble et al. 2008) and the current study (closed circle). The solid regression line in the center represents data including previous studies and the current study (log Cb = 1.0648 Cs6 + 0.7425, r2 = 0.89), and the other two lines represent the coefficient of variation. The equation for the relationship between measurements by bioaccumulation testing and Tenax extraction in previous studies (excluding the data from the current study) (log Cb = 1.0238 Cs6 + 0.8355, r2 = 0.87) is also shown
Fig. 2

The relationship between the matrix-SPME fiber concentrations (ng/ml) to L. variegatus bioaccumulation testing (log Cb, ng/g lipid) using data from previous studies (open triangle) (You et al. 2006; You et al. 2007a; Trimble et al. 2008) and the current study (closed circle). The solid regression line in the center represents data including previous studies and the current study (log Cb = 1.171 Cf + 0.3715, r2 = 0.86), and the other two lines represent the coefficient of variation. The equation for the relationship between measurements by bioaccumulation testing and matrix-SPME in previous studies (excluding the data from the current study) (log Cb = 1.0993 Cf + 0.501, r2 = 0.92) is also shown

As shown in Fig. 1, the 6-h Tenax extraction data predicted concentrations of HOCs accumulated in the worms at steady state very well, with a slope being near 1 and an r2 value of 0.89. Conversely, the SPME fiber concentrations slightly underestimated HOC body residues in the current study compared with previous studies (You et al. 2006, 2007a; Trimble et al. 2008). As previously stated, the analytical difficulty of measuring low concentrations of HOCs in the fibers caused high SDs in the matrix-SPME measurements, and this may partially explain the deviation noted in estimating the body residues. In addition, uptake of the HOCs by sediment ingestion may also have contributed to the chemical concentrations measured in the organisms compared with those estimated by matrix-SPME measurements, especially for HOCs in fine sediment. By combining the data from previous studies with those from the current study, it can be observed that Tenax extractions and matrix-SPME were able to predict concentrations in L. variegatus within a factor of two 79.9% and 42.4% of the time, respectively, and within a factor of four 95.4% and 85.9% of the time, respectively. Overall, results from the current study, similar to those from past studies, suggest that Tenax extraction and matrix-SPME are viable options as surrogates for measuring bioavailability of various HOCs. However, matrix-SPME may underestimate bioaccumulation potential due to sediment ingestion acting as a second uptake route.


The bioaccumulation potential and kinetic rates were compared for HOCs in sediments with different particle-size distributions. Although no difference was noted in BSAF values for HOCs in different sediments, toxicokinetic rates to organisms and chemical concentrations in sediment porewater varied. The key, however, is determining if these subtle toxicokinetic differences noted during the bioaccumulation testing are in fact biologically relevant. More thorough studies evaluating more distinct particle-size distributions and more potent contaminants can better evaluate the biological relevance of the difference in toxicokinetic rates and their impact on bioavailability and toxicity. In addition, differences in toxicokinetic rates among sediments suggest that the role of sediment ingestion played in organism uptake should not be overlooked, especially for fine-grained sediments.


This work was partially supported by the Hundred Talents Program of the Chinese Academy of Sciences (kzcx2-yw-BR-05) and the National Natural Science Foundation of China (40971263). The authors thank Philipp Mayer for providing the SPME fibers. GIGCAS contribution Number 1252.

Copyright information

© Springer Science+Business Media, LLC 2010