DECIDING WHEN TO INTERVENE

Managing contaminated sediment usually requires a balance between environmental and economic considerations. Depending on one's perspective, there are environmental goals in conflict with economic realities, or there are economic goals in conflict with environmental realities. These two viewpoints are really opposing sides of the same coin. Depending on the applicable goals, practices, or regulations, managers can employ many different decision criteria for evaluating site remediation/restoration, source control, or dredged material disposal options. For example, one could maximize the probability of achieving environmental goals, minimize the cost within acceptable risk boundaries, or optimize cost/benefit. Regardless of particular decision criteria, there are three critical elements for effective management of contaminated sediment: risk-based decision-making, weight of evidence assessment, and resource condition monitoring. These elements are dependant on one another (i.e., decisions require assessment, assessment requires monitoring).

The central tenet of a weight of evidence approach is that multiple lines of evidence should support decision-making. The corollary is that no single line of evidence should drive decision-making (unless you believe that a single line of evidence gives you all the information necessary, and you are willing to accept the outcome). A weight of evidence assessment can be implemented in a tiered fashion, with increasingly complex evaluations undertaken only as needed to reduce uncertainty (Ingersoll et al. 1997). In a tiered approach, the weight of evidence required should be proportional to the weight (e.g., cost) of the decision.

Contaminants in sediment can cause adverse effects either through direct toxicity to benthic organisms or through bioaccumulation and food chain transfer to human and wildlife consumers of fish and shellfish. Sediment quality assessments are best performed using a weight of evidence approach that incorporate sediment chemistry, laboratory studies of toxicity or bioaccumulation, and field evaluation of the benthic community or fish tissue residues. These lines of evidence can be organized into a sediment quality "triad" that provides the framework for these assessments (Long and Chapman 1985). "Triads" exist for evaluating risk to benthic organisms and risk to human and wildlife consumers.

The problem formulation step of risk assessment involves the a priori identification of assessment endpoints (i.e., what is to be protected) and measurement endpoints (i.e., what lines of evidence to evaluate). For a sediment ecological risk assessment, the assessment endpoint may be "a healthy benthic community free from contaminant-induced degradation". Sediment quality triad measurement endpoints are sediment chemistry, sediment toxicity, and benthic community condition. Each of these are associated with uncertainties as they relate to the assessment endpoint. For example, sediment chemistry data do not demonstrate whether measured contaminants are bioavailable. Sediment toxicity tests can indicate bioavailability of contaminants, but test conditions may not reflect field conditions. In addition, the test organisms may not adequately reflect the sensitivity of the full range of species comprising the benthic community. Benthic community condition may reflect degradation from factors other than contaminants (e.g., low dissolved oxygen). Reliance on any one of these measurement endpoints to evaluate exposure and effects is problematic for characterizing risk. In contrast, a weight of evidence assessment using all three gives the assessor much more information to reach conclusions.

Each presenter at this workshop was asked to describe an assessment approach suitable to support a decision "to act" or "take no further action", with a focus on the scientific tools used in each approach. The answer to the question depends greatly on the action contemplated, and a tiered approach may be most appropriate. However, the "yes/no" aspect to the question is analogous to asking whether or not applicable water quality standards (or objectives) are met. Citizens, responsible parties, and regulated entities expect consistency in the approaches used to assess the condition of protected resources. A good way to achieve this consistency is through State/Provincial adoption of sediment quality standards/objectives and supporting implementation procedures. In the U.S., a water (or sediment) quality standard includes a designated use for the waterbody (i.e., the assessment endpoint) and criteria to meet the designated use (i.e., the measurement endpoints). Criteria can either be numeric ("10 ppm") or narrative ("no toxics in toxic amounts"). Under the U.S. EPA's current policy of independent application, use of numeric criteria conflicts with a weight of evidence approach. However, a narrative criterion can accommodate a weight of evidence approach by design, which can be specified in implementation procedures.

A general model for sediment quality standards for the protection of benthic organisms is the focus of the remainder of this paper. The model is a tiered approach using sediment quality guidelines to evaluate sediment chemistry data as a first step. If guidelines are exceeded, and the weight of the decision requires reducing the uncertainty associated with either contaminant bioavailability or possible effects caused by unmeasured contaminants, standard sediment toxicity tests are used to determine if the standard is met. In this approach, the uncertainty surrounding contaminant bioavailability outweighs the uncertainty in test species sensitivity. This framework is consistent with EPA's current thinking, as stated in EPA's contaminated sediment management strategy: "EPA intends to encourage the States to use biological sediment test methods and sediment quality [guidelines] to interpret the narrative standard of 'no toxics in toxic amounts'" (USEPA 1998).

Sediment chemistry provides information on contaminant concentrations and related chemical variables. Sediment quality guidelines (SQGs) help determine whether contaminants are present in amounts that could cause or contribute to adverse effects. Guidelines based on equilibrium partitioning address bioavailability and can set protective levels for specific contaminants. DiToro et al. (1991) describe the technical basis for deriving guidelines using equilibrium partitioning (EqP) theory.

Equilibrium partitioning-based sediment guidelines (ESGs) are based on the theory that an equilibria exists among contaminant concentration in sediment porewater, contaminant attached to a binding phase in sediment (e.g., organic carbon, sulfide), and biota. ESGs are derived by assigning a protective water-only effects concentration to the porewater (such as an FCV), measuring the principle binding phase for a particular contaminant (e.g, fraction of organic carbon for nonionic organics, acid volatile sulfides for metals), and applying a contaminant specific partition coefficient if necessary (e.g., K_oc). For nonionic organics, supporting laboratory spiked sediment data show that the predicted sediment toxicity units (based on LC50) agree with percent amphipod mortality within approximately a factor of 2. For metals, spiked sediment are not toxic to amphipods if the molar concentration of simultaneously extracted metals (SEM) does not exceed the molar concentration of acid volatile sulfides (AVS).

Contaminants almost always occur as mixtures in sediment. Guidelines for chemical mixtures are most useful for contaminants that tend to co-occur, have the some toxic mode of action, and have the same factors control bioavailability. Field data from sites thought to be exclusively contaminated with mixtures of PAHs (Swartz et al. 1995) effectively illustrate what sediment chemistry data can tell you and what it cannot tell you. Although sediment quality guidelines bound the range of tested amphipod mortality (i.e., provide a good screen), the range of concentrations where toxicity may or may not occur exceeds an order of magnitude. It is within this range that sediment toxicity tests can help determine whether effects are occurring and if standards are met.

Sediment toxicity tests provide a direct measure of effects, account for bioavailability, and can be standardized for multi-region use. On the other hand, they may not adequately represent all species that a standard intends to protect, they cannot differentiate among contaminant or natural geochemical causes of toxicity, and they may not represent field conditions (e.g., sediment collection, handling, storage, and manipulation may alter the natural bioavailability). In situ application of toxicity tests can help mitigate the latter limitation.

Several EPA and ASTM standard methods are available. EPA methods include short-term and long-term exposure tests for survival, growth, and reproduction of freshwater midge larvae and freshwater and marine amphipods. Use of standard methods increases data accuracy and precision, facilitates test replication, increases the comparative value of test results, and ultimately, increases the efficiency of regulatory processes requiring sediment tests.

The full sediment quality triad includes assessment of the benthic community condition. The methods available are many and varied, ranging from simple presence of indicator organisms, to areal abundance of species, to complex multi-metric statistical indices. Some of these methods are currently in use in various State water programs (e.g., Ohio), and could be specified in narrative standard implementation procedures as an additional tier. However, benthic community assessment is not a predictive tool: the effects have already been manifested. Elevated levels of contaminants in sediment, along with demonstrated laboratory or in situ toxicity, may be sufficient to "take action" to prevent degradation even in cases where a benthic community assessment does not indicate impairment.

Available field data indicate that sediment toxicity tests are predictive of benthic community impairment. A recent analysis of regional sampling data indicated that reduced amphipod survival is predictive of benthic community degradation approximately 75 percent of the time (Scott 1998). An example of this relationship using samples taken from Baltimore Harbor (McGee et al. in review) shows that a line drawn at 70 percent test survival effectively divides amphipod abundance into distinct groupings: samples with fewer than 100 organisms per square meter and samples with greater than 100 organisms per square meter. However, other field data sets suggest that survival alone may not be sufficiently protective or predictive of field conditions. Samples from coastal Southern California indicate that moderate degradation, as measured by amphipod abundance and species richness, occurs co-incident with moderate chemical contamination, yet without associated reduction in amphipod survival from laboratory toxicity tests (Swartz et al. 1986). The implication that sub-lethal effects may be responsible for field population effects is supported by laboratory data measuring multiple test endpoints. Using dose-response curves for amphipod survival, growth, and reproduction, moderate levels of contamination (i.e., 10% of a highly toxic sediment), there is no reduction in survival; yet reduction by half in growth rate and by a factor of 4 in reproduction as measured by offspring per female (DeWitt et al. 1997).

Considering the above information, an example implementation procedure for a narrative sediment quality sediment is provided. This freshwater example makes use of protective ESGs and predictive standardized toxicity tests using a lethal and sub-lethal endpoint for two representative species. Key elements of a successful sediment management program include frequent monitoring of resource conditions; field evaluation of sediment chemistry, biological tests, and benthic community; harmonization of affected regulatory programs; and agreement among stakeholders and scientific peer review. One of these key elements, harmonization of affected regulatory programs, is illustrated by the sediment management approach in Washington State. In Washington State, source control, dredged material disposal, and contaminated sediment cleanup programs all share the same sediment quality standards. These standards are implemented using sediment quality guidelines and sediment toxicity tests, and this Washington program serves as the model for the framework proposed in this paper (Ecology 1995).

References

DeWitt, T. H., Pinza, M. R., Niewolny, L. A., Cullinan, V. I. and B. D. Gruendell. 1997. Development and Evaluation of a Standard Marine/Estuarine Chronic Sediment Toxicity Method Using Leptocheirus plumulosus. Draft Final Report prepared for U.S. EPA under contract DE-AC06-76RLO 1830 by Battelle Marine Sciences Laboratory, Richland, WA.

DiToro, D. M., Zarba, C. S., Hansen, D. J., Berry, W. J., Swartz, R. C., Cowan, C. E., Pavlou, S. P., Allen, H. E., Thomas, N. A. and P. R. Paquin. 1991. "Technical Basis for Establishing Sediment Quality Criteria for Nonionic Organic Chemicals Using Equilibrium Partitioning." Environmental Toxicology Chemistry. 10:1541-1583.

Ecology. 1995. Washington State Department of Ecology. Sediment Management Standards. Chapter 173-204 WAC. December 1995.

Ingersoll, C. G., Dillon, T. and R. G. Biddinger (eds.). 1997. Ecological Risk Assessment of Contaminated Sediment. SETAC press: Pensacola, FL. 189 pp.

Long, E. R. and P. M. Chapman. 1985. A Sediment Quality Triad: Measures of Sediment Contamination, Toxicity, and Infaunal Community Composition in Puget Sound. Marine Pollution Bulletin. 16: 405-415.

McGee, B. L., Fisher, D. J., Yonkos, L. T., Ziegler, G. P. and S. Turley. In review. "Assessment of Acute Sediment Toxicity in Baltimore Harbor Using the Estuarine Amphipod, Leptocheirus plumulosus." Environmental Toxicology Chemistry.

Scott, K. J. 1998. Relationship between Ampelisca sediment toxicity and parameters of benthic community condition. Presentation at the Society of Environmental Toxicology and Chemistry 19th Annual Meeting, 18 November 1998.

Swartz, R. C., Schultz, D. W., Ozretich, R. J., Lamberson, J. O., Cole, F. A., DeWitt, T. H., Redmond, M. S. and S. P. Ferraro. 1995. "EPAH: A Model to Predict the Toxicity of Polyaromatic Nuclear Hydrocarbons in Field-collected Sediment." Environmental Toxicology Chemistry. 14: 1977-1987.

Swartz, R. C., Cole, F. A., Schultz, D. W. and W. A. DeBen. 1986. "Ecological Changes on the Palos Verdes Shelf Near a Large Sewage Outfall: 1980-1983." Marine Ecology: Progress Series. 31: 1-13.

USEPA. 1998. EPA's Contaminated Sediment Management Strategy. EPA-823-R-98-001. U.S. Environmental Protection Agency, Office of Water. Washington, DC.