SUMMARY OF THE REMEDIATION TECHNOLOGIES DEVELOPMENT FORUM
IINERT SOIL-METALS ACTION TEAM
CONFERENCE CALL

2:00 p.m.–3:00 p.m.
August 19, 1998

Note: Items in red represent information that remained in question at the end of the conference call and will be discussed further in a future call.

On Wednesday, August 19, 1998, the following members of the IINERT Soil-Metals Action Team met in a conference call:

Bill Berti, DuPont Life Sciences (Co-chair)
James Ryan, U.S. Environmental Protection Agency (EPA) (Co-chair)
Sally Brown, U.S. Department of Agriculture (USDA)
Rufus Chaney, USDA
Mark Doolan, EPA
Andrew Green, International Lead Zinc Research Organization (ILZRO)
David Mosby, Missouri Department of Natural Resources

Also present was Christine Hartnett of Eastern Research Group, Inc. (ERG).

JOPLIN DATA

The conference call participants discussed animal-dosing, plant, in vitro, and soil data collected from the Joplin site. Participants also discussed data discrepancies and future activities.

Animal-Dosing Tests

Sally Brown noted that USDA performed another rat-dosing study. This study, she said, utilized laboratory incubated samples. Trying to make a comparison with the University of Missouri’s pig-dosing studies, Brown continued, USDA used some of the University’s soil treatments and dosed rats three doses rather than just one. Brown said results are not yet available.

Pig-Dosing Tests

A recent pig-dosing study, David Mosby noted, generated some confusing results. In the 0.5% phosphorus (0.5% P) incubated sample, Mosby continued, lead bioavailability was higher than in the control. According to Bob Blanchar, Mosby reported, this strange result may be attributed to numerous redox reactions. Each time the sample went through a wet and dry cycle, Mosby said, iron switched from a reduced to an oxidized form. (Iron was not added to the treated sample. The iron discussed is that which naturally occurred in the soil.) As iron reoxidizes, Mosby explained, it precipitates out. When the pH is low, he continued, iron plus phosphate precipitates out. In so doing, the reaction essentially starves the treatment plot of phosphate.

Bill Berti said pig-dosing studies suggest that lead bioavailability is reduced by 40% in the 0.5% P and 1% P field treatments. Mosby said his calculations differ somewhat, indicating a 26% and 31% reduction with the 0.5% P and the 1% P treatments, respectively.

Plant Data

Collecting Samples and Distributing Data

Brown noted that plant samples were collected in September 1997 and May 1998. To avoid including soil in the plant sample, samples were collected about 1 inch above the ground. To further ensure that plant tissues were soil-free, Brown said, they were washed at the USDA laboratory. Brown reported that she has generated plant data for the September 1997 and May 1998 sampling events. She agreed to distribute these data to the conference call participants via Berti.

Mosby reported that he and John Yang may collect additional plant samples later in the year. Mosby said he is willing to collect extra plant samples if anyone needs them. He asked how people would want the samples shipped. Brown noted that the May 1998 plant samples were sent via UPS. Even though they were not refrigerated, she continued, the samples were still fresh enough to withstand a washing after they arrived at the USDA laboratory.

Lead Concentrations and Reductions in Lead Bioavailability

Rufus Chaney noted that lead concentrations varied considerably between the September 1997 and May 1998 plant samples. In September, lead concentrations were 23.3 parts per million (ppm) and 4.5 ppm in the control and phosphorus-treated samples, respectively. In May, the control samples had 5.7 ppm and the phosphorus-treated samples had 1.27 ppm. Chaney noted that the September 1997 plant samples were collected from a newly tilled area where most of the plants were young. By May 1998, he continued, the plants were older and better established. Berti asked if there were theories to explain why the May 1998 samples had lower lead concentrations. Chaney said a number of variables have been postulated to influence plant uptake, including automotive exhaust, stagnant air during wintertime, and increased deposition. Also, he noted, one study shows that grasses have a higher lead concentration as autumn approaches. Given that the May 1998 plant samples were a mature and fully developed stand of grass, Chaney said that he suspects the May samples reflect the long-term normal pattern at the Joplin site. Mark Doolan asked whether meaningful differences between controls and treated plots can be obtained when lead concentrations are so low. Chaney and Brown said that they can. Doolan felt more confident about the data after being told that the samples were analyzed using an Inductively Coupled Plasma-Atomic Emission Spectometer (ICP-AES).

Berti said he has started summarizing some of the September 1997 and May 1998 plant data. As he indicated in a recent memo (dated August 19), Berti’s evaluation indicated that the 3.2 % P, triple phosphorus super (3.2% P, TSP) treatment caused an 80% reduction in plant lead availability. Berti estimated bioavailability reduction by calculating:

([Average amount of plant lead in control plots] - [Average amount of plant lead in 3.2% P TSP plots])
[Average amount of plant lead in control plots] * 100

Cadmium and Zinc Reductions

Brown and Chaney noted that reductions in cadmium and zinc were observed in the plant data. Concentrations across controls and treatments differed as follows:

Chaney noted that lime was added to the phosphorus treatment, raising the pH from 6.45 to 6.74. Doolan doubted that the change in pH accounted for the reduction in cadmium and zinc bioavailability.

Biomass Considerations

Berti asked whether Brown measured the biomass of the harvested plants. Brown had not. Berti said some people have been asking whether there is much difference in biomass production between control and treatment plots.

In Vitro Test

Brown noted that soil samples were collected from the Joplin site in September 1997 and May 1998. In vitro tests were performed on both sets of samples. The September 1997 samples were analyzed by USDA, using a glycine buffer and a pH of 2.2. The May 1998 samples are being analyzed by Dr. John Drexler using the Solubility Bioavailability Research Consortium (SBRC) protocol. The SBRC protocol calls for a pH of 1.5, end-over-end shaking at 30 revolutions per minute, 0.4 molar glycine buffer, and 37·C. The SBRC methodology, Berti noted, is being tested in numerous laboratories. During Phase I of the testing, the same sample was sent to three laboratories to determine whether results are reproducible. According to Berti, the results were reasonably consistent. Now, Phase II of the testing has been initiated. Berti said the SBRC test in vitro data have not yet been correlated with in vivo data for treated soils. Berti said in vitro data cannot yet be correlated with results obtained from pig-dosing studies because not enough different samples have been run through both models. Doolan noted that some people do not think the data correlate.

Brown has started analyzing Drexler’s in vitro data for the May 1998 samples. She compared lead bioavailability reductions as estimated using Drexler’s in vitro test (the May 1998 samples), the USDA in vitro test (the September 1997 samples), rat-dosing data, and plant data. She said Drexler’s data did not agree with data generated from the rat-dosing and plant studies. Specifically, she continued, lead bioavailability reduction was more pronounced in the rat-dosing study than in the in vitro test. Doolan was puzzled by the result, noting that the opposite was shown when results of in vitro tests were compared with those from pig-dosing studies.

Berti asked whether Drexler’s data showed lead bioavailability differences between treated samples and control samples. Brown said the only pronounced difference was observed in the 2.5 Iron, 1% P treatment. With this treatment, she said, lead was less bioavailable than in the control. Brown said the opposite was observed in the 1% P field sample; that is, a 0.05% increase in bioavailability was observed for the 1% P treatment. Brown said differences between the control and the 10% compost treatment were very small, with the treated sample showing only a 0.5% reduction in lead bioavailability. Such a small reduction correlates poorly, she said, with the 36% reduction that was observed in one rat-dosing study. Berti noted that these findings are consistent with those reported by Gary Pierzynski. Berti said that Pierzynski and a graduate student ran the in vitro test using 0.4 molar glycine and a pH of 1.5. They too, Berti continued, failed to see significant solubility changes across treatments.

Berti said the IINERT group needs to talk more about why the SBRC in vitro test fails to detect reductions in treated samples. The reductions are apparent in plant and animal studies, he stressed, but are not being captured in the in vitro studies. Brown noted that better correlation was obtained when USDA performed the in vitro test at a higher pH (2.2) than that recommended by the SBRC test (pH of 1.5).

Soil Data

Berti said his memo summarizes total analysis data for the May 1998 samples. Berti said total phosphorus analysis indicates that the control had about 2,000 ppm phosphorus. When the 1% P, TSP; the 1% P (as rock phosphate); and the 3.2 % P samples were analyzed, results indicated that they were closer to 0.8 % P, 1.2% P, and 2.8% P, respectively.

Discrepancies

Some data discrepancies have already been listed in the discussion above. One discrepancy that was particularly puzzling to Brown was the fact that the 1% P (as phosphoric acid) treatment did not perform better. She noted that Mosby’s 1% P (as phosphoric acid) treatment appeared to be more effective than the 1% P, TSP treatment and asked whether the difference could be related to pH. (As noted by Brown, the pH is brought down lower in the 1% P treatment than in the 1% P, TSP treatment.) Mosby said that Blanchar believes the pH has to be lowered and then brought right back up. Mosby said his team waited about a week before placing lime on the samples. Mosby and Doolan noted that there is a lag time for the pH to be brought back up.

Future Activities

Mosby said his team has about $100,000 left in its grant. The team plans to use this money to:

• Collect another round of soil samples. Mosby said he wants to collect new soil samples and run them through the pigs.
• Rerun the 1% P pig-dosing study. Mosby said he wants to dose his pigs with the 1% P treatment that has been sitting in the field for a year. Running this study, he said, would help answer the question about what happens over time.
• Evaluate the 0.5% P treatment plots in low lead concentration soils. Mosby noted that Joplin is a Superfund site and needs to be remediated. If IINERT technology is going to be used, he noted, it will be used on areas with low levels of contamination.

Brown asked Mosby to wait until the IINERT group has a chance to meet in person (see below) before collecting data. Mosby said he would do so if possible, but noted that his schedule is dictated by other people.

NEXT MEETING

James Ryan said the IINERT RTDF group will likely meet as a whole around the beginning of 1999. Conference call participants agreed that a small group needs to meet prior to the general meeting. Such a meeting, Ryan noted, will allow investigators to review and analyze all available data and could help generate an explanation for some of the strange results that have been coming forth.

At this point, participants agreed that the meeting should be restricted to those who have data to present. Berti said that he would draft a list of 10 to 12 possible attendees and distribute it to the conference call participants for discussion. The group agreed that Stan Casteel, Blanchar, and John Yang should be in attendance. To accommodate them, the group decided to meet in Kansas City. The date for the gathering has been tentatively set for October 7 and 8, 1998. By this time, Casteel and Mosby are expected to have additional reports completed.

MISCELLANEOUS

Mosby said that his group lost several root and plant tissue samples due to a malfunction in the refrigerator system.

Berti asked Mosby to send his plot numbers. Berti said that this was the only information missing from the data spreadsheet that he is generating.

NEXT CONFERENCE CALL

The next conference call is scheduled for September 16, 1998, between 2:00 and 3:00 p.m. eastern daylight time.