DECIDING WHEN TO INTERVENE

Summary

Toxicity evaluation under controlled laboratory conditions is an important component of sediment risk assessment. It is commonly accepted that a single species can never adequately reflect contaminant effects to all biota in the aquatic ecosystem under study. This calls for the use of several test species representing different trophic levels in the test battery. Micro-scale toxicity tests have shown good correlation with macroinvertebrate assays, benthic organism responses, and contaminant levels. These tests are proving to be attractive to the scientific community at large because of their cost-effectiveness in providing rapid and reliable results. However, the use of a testing battery requires a tool to integrate multiple toxicity data in a ranking process that allows the managers to determine the extent of the problem, identify hot spots and assess the need to "act" or "take no further action" on a particular site. The goals of this paper are to introduce the Sediment Toxicity (SED-TOX) Index for the assessment and ranking of toxic hazards in sediment, illustrate its application with results from a battery of bioassays performed on four exposure phases (pore water, wet sediment, organic extract, and whole sediment), and compare the SED-TOX scores with benthic community metrics. The discussion will emphasize on how the Index can be used to make sediment management decisions.

Introduction

Contaminated sediment samples may contain complex mixtures of contaminants. In such cases, it is widely accepted that one cannot rely on a single bioassay to detect all potential hazards (Cairns 1986; 1988). If the purpose of toxicity testing is to protect the environment from the action of toxicants, the testing program must optimize its ability to detect contamination. Batteries of tests are now commonly used for that purpose. The assessment of toxicity in multiple bioassays provide data that may be used to assess integrated responses at several levels of organization simultaneously. A battery of tests typically covers several trophic levels and several effect endpoints (e.g., enzymatic activity, genotoxicity, growth, reproduction, survival). The battery approach for the assessment of contaminated sediment has been recommended by several organizations such as Environment Canada, the US. EPA, and the International Joint Commission.

Test batteries may, however, generate contradictory results in the data set, which may lead to difficulties in the decision-making process related to the management of contaminated sediment. This complexity calls for a mathematical tool to integrate toxicity data and provide comparable indices for comparing test sites with reference sites, identifying hot spots, determining spatial gradients to identify contamination source, monitoring following remedial actions, and establishing criteria. Between 1994 and 1998, efforts have been devoted at the St. Lawrence Center to develop an index which could integrate multiple endpoint measures into a single value. The resulting SED-TOX Index evaluates and compares the relative hazard associated with contaminated sediment from different sites based on a suite of toxicity tests (Bombardier and Bermingham 1999). Following its development, we have expanded upon the SED-TOX approach in efforts to develop a basis for evaluating benthic impacts in relation to multiple measures of toxic responses. This presentation will describe the SED-TOX Index, provide an example for its application, and show how SED-TOX scores can be used to predict benthos degradation.

Some Terminology

Hazard: Likelihood of adverse toxic effects occurring as a result of exposure to one or more contaminants at a particular site.

Battery approach: The use of a variety of species representative of different trophic levels and sensitivity to toxicants to evaluate the toxic potential of contaminant mixtures, considering several exposure routes. It is believed that using several test organisms and exposure phases increases the probability of correctly identifying sediment that would be expected to be toxic to aquatic organisms under field conditions.

Exposure phase: The matrix used in the sediment toxicity tests (e.g., pore water, elutriate, unmodified whole sediment). Tests on pore water, which typically contains free salts, solutes, colloidal material, and/or organic solutes, provided information on the toxicity of dissolved substances in the aqueous phase. Wet sediment phase tests yielded information on the potential toxic effects of contaminants sorbed to sediment particles that could be released in the water column during disposal of dredged material or during resuspension events. Whole sediment bioassays measured the effect of all bioavailable contaminants, where bioavailable is defined as the fraction of the total contamination in the interstitial water and sediment particles that is available to aquatic organisms. Finally, exposing organisms to organic extracts constituted a worst-case scenario since extracting sediment with a solvent such as methylene chloride releases toxic molecules which become much more bioavailable in toxicity testing. "Worst-case" implies that the effects demonstrated with sediment organic extracts may never be demonstrated in the field, but if they are, this type of testing can be considered proactive.

Recommendation of an appropriate testing strategy for the assessment of sediment - Environment Canada (Quebec Region)

In 1995, a joint venture partnership was struck between BEAK International and Environment Canada (St. Lawrence Centre, Eco-Innovation Technologique) to conduct a comprehensive study to assess the suitability of various microscale tests for sediment toxicity assessment, and to recommend an appropriate battery for freshwater sediment. The performance of microscale tests was appraised by comparing their responses with those of macroinvertebrate assays, benthic community structure indices, and sediment contaminant characteristics (Côté et al. 1998a; 1998b). Twenty different toxicity test methods were performed on 15 sediment samples and evaluated for their inclusion in the testing strategy:

Pore water (a):

Bacteria:

• Microtox acute test
• Mutatox
• SOS Chromotest
• Microtox microplate
• Microtox chronic test

Algae:

• Chronic test
• Acute test

Microinvertebrates:

• Thamnotoxkit
• Rotoxkit
• Daphtoxkit
• Algaltoxkit
• Hydra

Pore water (b):

Biomarkers:

• Hepatocytes (viability, MFO, MT, DNA damage)

Wet/Whole sediment (c):

Bacteria:

• Microtox solid phase test
• ToxiChromoPad

Algae:

• Direct contact test

Macroinvertebrates:

• C. riparius (survival, growth)
• H. azteca (survival, growth)

Organic extract (d):

Bacteria:

• Microtox acute test
• Mutatox
• SOS Chromotest
• Microtox microplate
• Mictotox chronic test

Microinvertebrates:

• Thamnotoxkit
• Daphnia IQ
• Hydra

n = 21 biotests

After assigning hazard categories for each bioassay, response agreement of microscale bioassays was gauged against macroinvertebrate bioassays, contaminant levels, and measures of benthos degradation. Tests which showed good concordance with chemistry, macroinvertebrate test results, and benthic community data were retained for their inclusion in the battery approach, along with two conventional bioassays:

Trophic level	Assay	Exposure phase
Primary producers	Selenastrum capricornutum (Direct contact test)	Wet sediment
Primary consumers	Hyalella azteca, Chironomus riparius (survival)	Whole sediment
	ThamnotoxkitTM, DaphtoxkitTM	Pore water
Secondary consumers	Hydra attenuata (tentacles morphology)	Pore water
Decomposers	Microtox - Vibrio fischeri	Pore water
	SOS Chromotest (genotoxicity) Escherichia coli PQ37	Pore water

SED-TOX calculation

Once toxicity has been assessed through the use of such a battery, data can be integrated in the SED-TOX formula as follows:

1. Data are assembled in the SED-TOX spreadsheet (Excel^TM)
2. Data are converted in TU adjusted for sediment dry weight
3. Mean toxic scores are derived for each exposure phase (WAPT scores - or weighed average phase toxicity)
4. Bioassays are weighed according to sensitivity
5. WAPT scores are cumulated, and divided by the total number of exposure phases considered in the battery (r), resulting in the CAPT (cumulated average of phase toxicity) score
6. CAPT scores are multiplied by the number of exposure phases that elicited toxicity, and then log transformed to result in the SED-TOX score

The formula for the Index calculation is as follows: SED-TOX = log10[1+n(CAPT)]

where: n = number of phases eliciting toxicity
CAPT = cumulative average of phase toxicity
and n(CAPT) is the Toxic Print

A spreadsheet program has been developed by the St. Lawrence Centre to perform those calculations automatically. This program is available for free.

Cutpoints separating four toxic hazard levels were defined. SED-TOX scores may vary from 0 to 4. A SED-TOX score of 0 indicates no toxic hazard potential; for values between 0.1 and 1, the toxic hazard is considered marginally toxic; for values between 1.1 and 2.0, toxic hazard is considered moderately toxic; and scores greater than 2 are considered highly hazardous.

Application

The Index was applied to data collected from 2 marine sectors in the Gulf of St. Lawrence. Three sites were assessed within each sector (a site considered for dredging and suspected for presenting high levels of contaminants, a site considered for disposal of the dredged material, and a reference site located nearby). Results clearly indicated that the SED-TOX Index discriminated the hazard potential of the dredging sites (some stations showing high scores), as compared to their respective reference sites (mostly marginal scores). SED-TOX scores were then compared to chemical concentrations. Sites with the highest levels of contaminants had the greatest ratio of high SED-TOX scores, while those with toxicant levels below the sediment quality criteria had the highest ratio of marginal SED-TOX scores. However, the relationship was not linear.

Comparison with benthic community metrics

We also wanted to verify if the SED-TOX Index could be used to predict adverse effects on the benthic community (measured via benthic community responses in field samples). Data on benthos degradation were derived from 15 sediment samples taken from different areas in the St. Lawrence River. The battery of bioassays put forward by BEAK International and EC (see table shown above) was used to assess sediment toxic potential. Benthic community data were used to calculate a variety of metrics, including the total number of taxa, the Shannon-Wiener Index, taxa richness, and number of intolerant or tolerant species. All graphs showed a consistent pattern: degraded benthos was associated with high SED-TOX scores (i.e., > 2.0). Indeed, total number of taxa, the Shannon-Wiener Index, and taxa richness all decreased with increasing SED-TOX scores. Oligochaetes (tubificids) accounted for 75% of the benthic species in sites that showed high SED-TOX scores, as compared to 20% in sites with moderate toxicity scores. The relative abundance of sensitive species (e.g., caddisflies) was greater at sites with lower SED-TOX scores. These preliminary results suggest that the cutpoint of 2 for identifying highly toxic sediment may be useful at predicting degraded benthos. This toxic hazard threshold, however, remains to be validated with a greater number of sediment samples.

Concluding remarks

The advantages of the Index:

• Is founded on generally accepted concepts and principles
• Allows the incorporation of an unlimited number of toxicity tests; however, the use of redundant information may overload the Index and reduce its discriminatory power
• Considers all possible exposure routes for aquatic organisms
• Takes into account the relative sensitivity of the exposure phases and that of the effect endpoints
• Can be used on data already gathered - no age limit for data set

The disadvantages of the Index:

• Requires professional judgement to interpret SED-TOX scores
• Time of sample collection and composition of the test battery should ideally be the same
• Does not take into account the variability of the tests, therefore requires the use of standardized methods

Post-hoc evaluation of the Index - refinement of the Index should focus on the:

• Influence of inconsistency of test selection and temporal scale of sampling on Index scores
• Validation of the range of SED-TOX scores and threshold levels with field data
• Inclusion of sediment volume in the formula, so as to compare not only Toxic Prints but also Toxic Loads

Possible applications of the Index:

• Assess the need "to act" or "take no further action"
• Determine the spatial gradient of toxic hazard
• Establish priorities - identify hot spots
• Evaluate the progress of remediation
• Define the composition of a test battery - identify redundant tests to reduce the analytical effort at subsequent sites

References

Bombardier, M. and N. Bermingham. 1999. "The SED-TOX Index: Toxicity-directed Management Tool to Assess and Rank Sediment Based on their Hazard- concept and Application." Environmental Toxicology Chemistry. 18(4).

Carins, J., Jr. 1988. "What Constitutes Field Validation of Predictions Based on Laboratory Evidence." Aquatic Toxicology and Hazard Assessment, ASTM Special Technical Publication 971. W. J. Adams, G. A. Chapman, and W. G. Landis, eds. American Society for Testing and Materials. Philadelphia, Pennsylvania. pp 361-368.

Cairns, J., Jr. 1986. "What is Meant by Validation of Predictions Based on Laboratory Toxicity Tests." Hydrobiologia. 137:271-278.

Cote, C., Blaise, C. Schroeder, J., Douville, M. and J. R. Michaud. 1998a. "Investigating the Adequacy of Selected Micro-scale Bioassays to Predict the Toxic Potential of Freshwater Sediment Through a Tier Process." Water Quality. Res. Journal. Canada. 33:253-277.

Cote, C., Blaise, C., Michaud, J. R., Menard, L., Trottier, S., Gagne, F., Riebel, P. and R. Lifshitz. 1998b. "Comparisons Between Micro-scale and Whole Sediment Assays for Freshwater Sediment Toxicity Assessment." Environmental Toxicology Water Quality. 13:93-110.