Lauma Jurkevics, CCC
Tom Johnson, POLB
Nick Buhbe, Ogden Environmental
Bill Paznokas, CDFG
Terri Ely, Corps
Michael Lyons, RWQCB
Steve Bay, SCCWRP
David Moore, MEC Analytical
Anthony Risko, Corps
Guangyu Wang, SMBRP
Steven John, EPA
Brian Ross, EPA (via conference call)
Jay Field, NOAA (via conference call)

Jay Field from NOAA’s office in Seattle joined the subcommittee by conference call to discuss his work on compiling a national sediment/toxicity database. This multi-agency collaborative effort is attempting to match sediment chemistry and toxicity databases. Data included are from multiple sources, including the Great Lakes program and state data (California, Florida, Washington, Minnesota). The project has been a several-year effort to assemble a high quality database that involves a tight data screening process.

The database has over 2000 data points for amphipods (Ampelisca, which is mostly from the east and gulf coasts, and Rhypoxinius, approximately half from California and half from Puget Sound). The approach used allows for consideration of geographic factors and utilizes individual chemistry data end points. A screening process was developed to screen out samples where it is unlikely that the target chemical was responsible for observed toxicity.

Logistic modeling provides the basis for determining goodness of fit. Actual modeling numerous samples with toxicity at a given concentration allows for estimating toxicity values. This approach allows for comparison of different end points for different chemicals. Comparison of the Puget Sound and the California Rhypoxinius data demonstrated that for 18 of 25 chemicals there was no regional effect, thereby allowing evaluation of whether data from one area is relevant to different areas.

The study effort compared base toxicity (percent of toxic samples) in Ampelisca (approximately 25%) and Rhypoxinius (approximately 60%). By looking at different data sets, the NOAA method can be used for both highly and less contaminated sediments.

EPA commented that in California, a lot of toxicity testing is conducted with Ampelisca, and -- while it skews toward the lower concentrations -- there is effects-based testing at higher concentrations. EPA commented that the CSTF was contemplating combining local data with other data; NOAA stated that its approach is set up to accept additional high quality data.

Regarding the screening process used in this approach, NOAA discussed that the procedure was designed to exclude toxic samples for which the observed toxicity was unlikely to have been associated with the chemical of concern. In comparing screened and unscreened data, differences between the distributions of the two data sets were found at lower concentrations, with fewer differences in distribution at the higher concentrations.

NOAA described the Logistic Model as having a typical ‘S’ shaped dose-response curve and is used to evaluate the relationships between contaminant concentration and adverse effects for selected contaminants. Model provides for identification of chemical concentration associated with a particular (for example, 50%) level of toxicity. More protective levels (lower toxicity) would correspond to lower chemical concentrations.

Plugging AETs into a specific chemical model would yield a wide range of concentrations. The contaminant-specific logistic models have been used to evaluate the range of protection afforded by other SQG and by ERL/ERM values. There appears to be a more narrow range of effects for ERL/ERMs than for amphipod derived AETs.

NOAA stated that for severity of response modeling, by comparing to unscreened data sets, there was an increased severity effect with increased toxicity. While the procedure uses acute toxicity for amphipods, additional end points are being evaluated. Some initial comparisons between species have been done, but standardized definitions for "toxic" and "nontoxic" may be more important than species differences. Ancillary factors (ammonia, grain size, TOC, etc.) are being considered as well.

This model can also accommodate benthic endpoints; NOAA is considering developing a list of endpoints for a wide range of approaches. There are many benthic data available, although there are some problems with standardizing these data. NOAA has a wish list for more chronic data (some freshwater data -- for growth, survival and reproduction) and would like to include Neanthes.

NOAA has the goal of producing a set of models for individual chemicals and individual endpoints. NOAA does not see this for regulatory uses, but rather as a screening tool, as they recognize the problems associated with assessing contamination by looking only at chemistry and not at biological testing as well.

NOAA’s recommendation to the CSTF would be to first look at a larger database than what might be available for Southern California alone, and to try and determine whether there is any reason to suspect that local conditions or data are going to any different from the larger database. That is, evaluate how consistent the local data are with the national database. The difficulty is in getting enough data with a sufficient spread of chemical concentrations -- at least 500 samples

would be needed to develop models. However, to compare local data to a national database, much less data would be necessary.

The subcommittee discussed next steps and the need to reassess goals. EPA will prepare a revised goal statement, literature list, as well as proposed workplan and data gap report. The subcommittee discussed that the primary goal should be to focus on continuing to assemble data and then tapping into an existing effort, such as NOAA’s, to set screening levels. Additional work could be necessary for setting higher sediment values -- for contaminated or confined aquatic disposal levels.

Next Sediment Subcommittee meeting -- July 22, 1999, 10am - 12 noon, at LA District Corps offices.