TECHNICAL ANNEXE
A0254 - Development of a Practical Application of Biological Tests to provide a Risk Assessment Protocol for the Assessment of Dredged Material Disposal.
Milestone 2b. Develop procedures for testing for bioaccumulation of hazardous substances.
INTRODUCTION
Bioaccumulation refers to the accumulation of contaminants in the tissues of organisms through any route, including respiration, ingestion, or direct contact with contaminated sediment or water. In the USA, the EPA requires that bioaccumulation be considered as part of their regulations for environmental evaluation of dredged material proposed for ocean disposal. This consideration involves predicting whether there will be a cause-and-effect relationship between an animal's presence in the area influenced by the dredged material and an environmentally important elevation of its tissue content or body burden of contaminants above that in similar animals not influenced by the disposal of the dredged material.
Bioaccumulation tests have also been proposed as one of the biological assessments in the third tier of the proposed risk assessment protocol for the UK (see project CSG15). The presence of one or more measured contaminants in test organism tissue(s) indicates their biological availability in the dredged material, and assesses the potential for long-term accumulation of contaminants in aquatic food webs to levels that might be harmful to consumers (including humans) without killing the intermediate organisms. Bioaccumulation tests will be used to gather further information on the potential long-term impacts of a dredged material if the results from Tier 2 assessments are insufficient to inform a weight of evidence decision. For example, acute whole sediment tests may produce negative results but the chemistry indicates levels of persistent organic pollutants which may have an environmental impact. A bioaccumulation test would then be conducted to assess the significance of biological uptake of contaminants of concern.
To this end, procedures for testing for bioaccumulation from natural sediments were developed. A small study was set up with the aims of (a) testing the experimental design i.e. does bioaccumulation take place, and (b) identifying suitable test species.
The selection criteria for the test species were: they must be benthic, readily available, ingest sediments (directly or indirectly), be sensitive but tolerate handling/laboratory conditions, tolerate a range of grain sizes and provide adequate biomass for chemical analyses. For the purposes of the experiment, the chosen contaminant group of concern was polychlorinated biphenyls (PCBs), which are persistent, known to bioaccumulate and exert chronic rather than short-term acute effects. The King’s Dock in Swansea is an area known to be highly contaminated with PCBs (Reed, 2001) and therefore this site was chosen from which to obtain sediment samples.
MATERIALS AND METHOD
Sediment Sampling
Sampling for sediments took place in January 2002 at King’s Dock in Swansea. Three sites were sampled:
Site 1: 5136’997N 354’993W
Site 2: 5137’038N 354’843W
Site 3: 5136’883N 355’224W
A minimum of 150 litres of sediment was collected from each site using a Day grab on board a chartered vessel and placed in 10 litre plastic buckets. The sediment was immediately transported to the laboratory and refrigerated until required for the experiment. An equal quantity of clean reference sediment was collected from Creeksea, Essex by hand and refrigerated.
Sediment Preparation
To initiate the test, sediments were placed in two replicate tanks (A and B) per site and thoroughly homogenised, to give a sediment depth of approximately 10 cm. Subsamples were immediately removed for chemical analyses (10 g dry weight as a minimum) in order to confirm that they contained high levels of the contaminants in question. Flowing seawater (~1 L/min, to a depth of ~25 cm) and aeration were introduced 4 days before addition of the test organisms to allow conditions in the tank to reach equilibrium.
Species Selection
The following organisms were chosen for assessment of their bioaccumulation potential:
- Arenicola marina (Lugworm): This burrowing, bulk sediment ingesting polychaete is a standard test organism for assessing the short-term acute toxicity of sediments.
- Nereis virens (Common or Estuary Ragworm):A deposit feeding and carnivorous polychaete, which builds U-shaped burrows. Commonly used in sediment bioaccumulation studies.
- Mytilus edulis (Common or Blue Mussel): Filter feeding, bivalve mollusc. Frequently used in bioaccumulation studies with water only exposures.
- Crangon crangon (Brown Shrimp): A crustacean amphipod, which feeds on small plants and animals present on the surface of sediment.
- Platichthys flesus (Flounder): Bottom-living flatfish which feeds on small invertebrates in sediment.
Experimental Design
A minimum of 10g wet weight of tissue was required for chemical analysis per species and time period, which determined the minimum number of animals that were added to the test tanks, taking into account potential losses due to mortality.
Lugworms were obtained from a local bait supplier and N. virens was purchased from Seabait Ltd, a commercial ragworm farm. One hundred lugworm and 50 ragworm were placed in each of the ‘A’ replicate tanks, which were separated into two halves by dividers to prevent the ragworms from consuming the lugworms during the exposure period.
Juvenile flounder (~10 cm length) were obtained from the Port Erin Marine Laboratory, Isle of Man and acclimated to laboratory conditions for three days before being added to each of the ‘B’ replicates (30 per tank). Brown shrimp were collected from the Crouch estuary and maintained in tanks with flowing seawater until the start of the test, when 105 were added to each 'B' replicate. Blue mussels were obtained from Fence Bay Fisheries on the west coast of Scotland. Thirty mussels were placed in polypropylene baskets, suspended just above the sediment, in each of the 'B' replicates.
On day 0, the appropriate numbers of animals were added to the tanks and sub-samples frozen for chemical analysis. Subsequently, subsamples of each species were collected on days 7, 14, 21 and 28 for chemical analysis of whole body homogenates. Physical conditions (temperature, pH, salinity and oxygen content) in the overlying water were monitored regularly (3 times weekly) to ensure a suitable environment was maintained.
After removal of the test organisms from the sediment on the sampling dates, it was necessary to empty the digestive tracts of the polychaetes and the mussels, since this is an easier procedure than excising the tract. Sediment in the guts may contain inert constituents and the PCBs in forms that do not become biologically available during passage through the digestive tract. Lugworms were allowed to depurate gut contents in clean sandy sediment for 24 hours. Ragworms and mussels were placed in separate aquaria with clean flowing seawater for 24 hours.
Chemical analysis
Sediment method
Samples were air dried in a controlled environment, ground, sieved to <2000µm, and mixed with anhydrous sodium sulphate. Samples were then soxhlet extracted for six hours using a 50:50 mix of n-hexane/acetone. Sulphur residues were removed by the addition of copper turnings to the extraction flask. A 50ml aliquot of the 100ml sample extract was cleaned up and fractionated by column chromatography using partially deactivated alumina and silica. CB#53 and CB#155 were added as reference internal standards, and the chlorobiphenyl residues determined by high resolution GC-ECD. Quality was assured by the parallel analysis of a certified reference material with each batch, along with a full method blank.
Biota method
All samples were defrosted and homogenised before chemical analysis for the ICES suite of 7 congeners using gas liquid chromatography.
Samples were mixed with anhydrous sodium sulphate, and placed in a freezer for a minimum of 24 hours. Samples were then defrosted, and soxhlet extracted for four hours using n-hexane. A 50ml aliquot of the 100ml extract was lipid tested to determine the maximum amount of sample that could be placed onto the alumina column. An aliquot of the extract, as determined by the lipid test, up to a maximum of 25ml was cleaned up and fractionated by column chromatography using partially deactivated alumina and silica. CB#53 and CB#155 were added as reference internal standards, and the chlorobiphenyl residues determined by high resolution GC-ECD. Quality was assured by the parallel analysis of a certified reference material with each batch, along with a full method blank.
RESULTS
Physical Readings
Physical readings taken three times weekly in each tank for the duration of the test did not vary significantly from those taken at t = 0, and all were within the acceptable limits for maintaining the physiological integrity of the test organisms.
Chemical Analysis of Sediments
Sediments from the three sites at Kings Dock Swansea and the control sediment from Creeksea were analysed for the ICES suite of 7 PCB congeners i.e. PCBs #28, #52, #101, #118, #153, #138, and #180 before the experiment was initiated.
The results of this analysis are presented in Table 1. Individual congener and total PCB concentrations were significantly higher at all three Swansea sites compared to the Creeksea sediment. Total PCB concentration ranged from 54 to 137 ppb dry weight, which is between 2 and 3 orders of magnitude greater than the current Action Level 2 used in the chemical assessment of dredged material in England and Wales. Concentrations in the control (Creeksea) sediment were below the limit of detection (0.2 ppb dry weight). Sites were ranked Site 3 > Site 2 > Site 1 in terms of mean concentrations of the two replicate tanks. The results therefore confirmed the presence of high levels of the contaminants under study and enabled the experiment to proceed.
Although sites A and B from each site were intended to be replicates, total PCB concentrations did differ quite significantly between the replicates at Sites 2 and 3.
Table 1. Concentration of individual PCB congeners of the ICES 7 suite and total ICES 7 in sediments (ppb dry weight) from Kings Dock Swansea and Creeksea (control).
Site / CB#28 / CB
#52 / CB
#101 / CB
#118 / CB
#153 / CB
#138 / CB
#180 / Total ICES 7
1A / 8.8 / 7.7 / 10 / 9.2 / 12 / 12 / 7 / 66.7
1B / 9.8 / 11 / 12 / 14 / 13 / 14 / 10 / 83.8
2A / 13 / 12 / 15 / 16 / 17 / 18 / 12 / 103
2B / 14 / 6.4 / 10 / 10 / 12 / 13 / 7.2 / 72.6
3A / 7.9 / 5.8 / 17 / 12 / 35 / 35 / 24 / 136.7
3B / 8.4 / 4.8 / 7.2 / 7.2 / 9.7 / 10 / 6.7 / 54
Creeksea / <0.2* / <0.2* / <0.2* / <0.2* / <0.2* / <0.2* / <0.2* / 0
* 0.2 ppb dry weight is the Limit of Detection (LOD)
However, this does not have any implications for the biota results since bioaccumulation factors are calculated using the initial sediment concentration in each individual tank. The reasons for this discrepancy between concentrations in tanks probably arises from the way the samples were taken at the dock and subsequently distributed to experimental tanks. At each site, a total of 15 x 10L capacity buckets were filled with sediment, with the contents of one Day grab filling one bucket, approximately. Hence, there may have been some variability in the PCB concentrations between grabs. On return to the lab, due to the practical difficulties inherent in firstly mixing the contents of 15 buckets and then transferring the homogenised sediment into two replicate tanks, the buckets were randomly allocated to either tank A or B, and the sediment from 7.5 buckets homogenised in each tank. It is therefore entirely plausible that one tank could receive sediments that had a higher concentration of one or more PCB congeners, resulting in higher total concentration. Table 1 shows that this was indeed the case at Site 3, with congeners 101, 153, 138 and 180 being much higher in replicate A than replicate B.
Chemical Analysis of Biological Tissue
The bioaccumulation test ended on 12th April 2002, after which chemical analysis of the biological tissues began. In order to pre-empt unnecessary analytical costs, it was decided that the t =28 day samples would be analysed first. If no PCBs were found in the tissues at the end of the experiment, then it was likely that the earlier sampling times would also be negative for PCBs, hence further analysis would not be necessary. If significantly elevated tissue concentrations of PCBs were determined in any of the samples then the experiment was successful. The remaining samples could be stored for future investigation of uptake kinetics, if deemed necessary.
T=0
At t=0, total PCB concentration was very low in all test species, ranging from 0 (not detectable) to 0.8 ppb wet weight (Table 2). Mussels had the highest body burden, which is not surprising given their widely recognised attribute of being efficient bioconcentrators of contaminants from the environment. Although the mussels were obtained from a known clean site, they nevertheless will have bioaccumulated PCBs from presumably low background concentrations in natural waters. Arenicola also had a higher tissue burden, probably a result of being wild-caught specimens which were exposed to low levels of PCBs in natural sediments. Flounder and Nereis were obtained from farms and therefore, as would be expected, had the lowest initial concentration of PCBs.
Although the limit of detection for each of the congeners is normally quoted as 1ppb, much lower values than this are actually analytically measurable, which is a necessary prerequisite when conducting a bioaccumulation study, otherwise meaningless data would be obtained. Low tissue concentrations such as those found at the beginning of a study can be accurately measured as well as small increases over time.
Table 2. Concentrations of PCBs at t=0 in each of the test organisms.
Sum ICES 7CB
#28 / CB
#52 / CB
#101 / CB
#118 / CB
#153 / CB
#138 / CB
#180 / ppm wet wt / ppb wet wt
Nereis / 0 / 0 / 0 / 0 / 0.000043 / 0.00006 / 0 / 0.000103 / 0.103
Mytilus / 0 / 0 / 0.0001 / 0.0001 / 0.00033 / 0.00023 / 0 / 0.000760 / 0.760
Flounder / 0 / 0 / 0 / 0 / 0 / 0 / 0 / 0 / 0
Shrimp / 0 / 0 / 0 / 0 / 0.0000012 / 0.000052 / 0.000063 / 0.000116 / 0.116
Arenicola / 0 / 0 / 0.000068 / 0.000072 / 0.00016 / 0.00013 / 0.000052 / 0.000482 / 0.482
In bioaccumulation studies, the contaminant concentrations in tissues are usually normalised to lipid content. Total PCB concentrations normalised to per gram of lipid (at t=0) are presented in Table 3.
Table 3. Lipid content of test organism whole body homogenates, and lipid-normalised total PCB concentrations, at t=0.
g lipid per g tissue / ICES 7(µg/g lipid)
Nereis / 0.016 / 0.00644
Mytilus / 0.008 / 0.095
Flounder / 0.274 / 0
Shrimp / 0.004 / 0.029
Arenicola / 0.006 / 0.08
Flounder had the highest lipid content and shrimp the lowest. The order of lipid normalised PCB concentration was mussels>Arenicola>shrimp>Nereis>flounder.
T= 28 days
The increase in test organism tissue PCB concentrations per se after 28 days exposure to the different sediments is presented in Figure 1. In the control sediment, three of the species (mussels, flounder and Nereis) showed a small increase in PCB body burden, up to 7ppb, whilst Arenicola and shrimp showed a loss from the tissues. Four of the five species had greater rates of uptake in the Swansea Dock sediments than the control. Shrimp did not bioaccumulate PCBs from any of the sediments, tissue concentrations at each of the three sites being lower than those measured at t=0.
Figure 1. PCB (ICES 7 suite) uptake per se in five benthic species after 28 days exposure to contaminated sediments.
Of those that showed a measurable increase in body burdens, flounder had the highest, with an uptake of 14.5, 10.9 and 9.8 ppb at sites 1, 2 and 3 respectively. This was followed by Arenicola (increase of 11.4, 8.5 and 6.7ppb), Nereis (increase of 7.4, 14 and 6ppb) and mussels (1.6, 2.8 and 0.7ppb). For Arenicola and flounder, the greatest increase was at site 1, followed by site 2 then site 3. For mussels and Nereis, the greatest increase occurred at site 2, followed by site 1 and site 3.
Tissue PCB concentration data were subsequently normalised to lipid content of the tissue homogenates. The results are shown in Figure 2. A different picture to the non-normalised data was obtained. Total PCB uptake, when expressed on a per gram of lipid basis, was by far the greatest in the lugworm Arenicola, with values of 3.9, 1.1 and 2.2 μg g-1 lipid for sites 1,2 and 3 respectively. Nereis, flounder and mussels had similar lipid-normalised uptake values, ranging from 0.64 (Nereis, site 2) to 0.05 μg g-1 lipid (mussels site 3). Shrimp, once more, had negligible or negative uptake values.
The Biota-Sediment Concentration Factor (BSAF) was calculated for each species at each site. The BSAFs are calculated using the equation:
BSAF = (Ct/L)/(Cs/TOC)
Where:
Ct = tissue contaminant concentration (µg/g tissue)
L = tissue lipid concentration (g/g tissue)
Cs = sediment contaminant concentration (µg/g sediment)
TOC = total organic carbon of sediment (g carbon/g sediment)
Figure 2. Lipid-normalised PCB concentration data for each of the test species, showing uptake (μg g-1 lipid) after 28 days exposure.
Since the organic carbon content of the sediments was not measured, a constant value of 1% was used in the calculation of BSAFs. The results are presented in Figure 3. These indicate that the highest BSAF occurred at Site 1 with Arenicola, with a value of almost 60. Sites 2 and 3 for this species were both above 10. For the other test species (excluding shrimp) BSAF values were between 2 and 6.
Figure 3. PCB Biota-Sediment Accumulation Factors (BSAFs) for each of the test species and sediments.
DISCUSSION
Hydrophobic organic contaminants can accumulate in benthic organisms either from the aqueous phase or via the diet. The higher the octanol-water partition coefficient (Kow) the greater the partitioning of the chemical to sediment and hence the more important the sediment ingestion exposure route becomes relative to exposure to interstitial water or overlying water (Lamoureux and Brownawell, 1999). Sediment ingestion has been shown to be a significant or primary route of uptake to several species of deposit feeding organisms exposed to hydrophobic organic contaminants with log Kow values ≥5.5 (Fowler et al., 1978). Bioaccumulation of non-essential xenobiotics is of concern both for its possible effect on the organism and for the contamination of higher trophic levels, including humans, that may occur.
The purpose of Milestone 2b of this project was to develop procedures for assessment of bioaccumulation of contaminants from dredged material, as part of the wider aim to develop a risk assessment protocol for dredged material. This type of biological assessment, together with short-term acute and longer-term chronic whole sediment bioassays, is a necessary requirement within such a decision-making framework in order to ensure that a full evaluation of potential impacts of a contaminated sediment can be made. The results of bioaccumulation tests are used to predict the potential for uptake of dredged-material contaminants by organisms (Biddinger and Gloss, 1984; Kay, 1984). Much of the information on bioaccumulation in the literature is derived from studies with chlorinated pesticides, PCBs and other non-polar, persistent organics, some of which accumulate to high levels in organisms (Spacie et al, 1995). Because of this emphasis, the process of bioaccumulation has often been considered important only for relatively non-toxic, poorly metabolised substances. However, it must be borne in mind that in actual fact, bioaccumulation is a pre-cursor to all chemical toxicity; without some degree of accumulation, toxic effects resulting from interactions at target site(s) cannot take place.