SYMPOSIA PAPER Published: 01 January 1990
STP18961S

Statistical Approaches to Screening Hazardous Waste Sites for Toxicity

Source

The kinds of questions bioassays can answer, as well as their advantages, are given. The ability to answer questions, such as “where is the waste?” or “is it toxic?” results from field studies based on twelve procedural steps for conducting remedial action studies. Simple or stratified random, systematic, and judgment sampling are presented as three possible paradigms for field sampling.

Bioassay results from two previously reported field studies illustrate how maps of toxicity can be prepared based on systematic sampling and show how cleanup decisions can be made using bioassay results based on few samples.

In the first study, logarithmically spaced soil samples (0–15 and 15–30 cm depths) were obtained along four parallel transects (90 m long and 15 m apart) at the Rocky Mountain Arsenal. A total of 72 soil samples (36 at each of two depths) were subjected to phytoassay using lettuce seeds; most samples were also subjected to Daphnia, Microtox, algal, earthworm, and lettuce root elongation bioassays. These latter bioassay results (exception earthworms) were inconclusive regarding toxicity, but allowed us to ignore several classes of compounds (i.e., water-soluble heavy metals, herbicides, and insecticides) single prior results [1] using pure chemicals showed depressed algal growth in the presence of these contaminants.

In order to depict the spatial pattern of observed seed mortality at each depth, we used kriging (a statistical technique developed for use in the mining industry) to produce contour maps. The results clearly showed that lettuce seed mortality was higher in the 15–30 cm fraction, that waste-trench soil was highly phytotoxic, and that toxicity decreased as a function of distance from the trench. In addition, we found that mortality contours produced by kriging could be useful in site cleanup decisions.

The second study used a series of water and sediment samples collected from a narrow stream adjacent to a wood treatment plant in Canton, Mississippi. Both creosote and pentachlorophenol were used for wood treatment. Sediment samples (15 cm) were collected every 20 m in the visibly contaminated zones. Based on sample linear interpolation of bioassay results, we found that different bioassays led to different conclusions regarding the toxicity of different areas, suggesting that contaminants other than creosote may have caused the observed toxicity. Moreover, chemical analysis was an inaccurate predictor of toxicity.

Author Information

Thomas, JM
Athey, LA
Skalski, JR
Price: $25.00
Contact Sales
Related
Reprints and Permissions
Reprints and copyright permissions can be requested through the
Copyright Clearance Center
Details
Developed by Committee: D34
Pages: 140–162
DOI: 10.1520/STP18961S
ISBN-EB: 978-0-8031-5123-9
ISBN-13: 978-0-8031-1293-3