SUMMARY OF THE REMEDIATION TECHNOLOGIES DEVELOPMENT FORUM
PHYTOREMEDIATION OF ORGANICS ACTION TEAM
TOTAL PETROLEUM HYDROCARBONS IN SOIL SUBGROUP
CONFERENCE CALL



April 15, 1998
11:00 a.m. to 12:30 p.m.



On Wednesday, April 15, 1998, members of the Remediation Technologies Development Forum (RTDF) Phytoremediation of Organics Action Team, Total Petroleum Hydrocarbons (TPH) in Soil Subgroup, met via a conference call. The following members participated:

Phil Sayre, U.S. Environmental Protection Agency (EPA) (Subgroup Co-Chair)
Steve Rock, EPA
Evelyn Drake, Exxon Research and Engineering
Peter Kulakow, Kansas State University
Royal Nadeau, EPA
Jim Brown, ReAct
David Tsao, Amoco Research Center
Mike Reynolds, U.S. Army Cold Regions
Charlene Owens, Exxon Product Research

Also present was Ben Carlisle of Eastern Research Group, Inc. (ERG).


DISCUSSION OF FIELD STUDY PROTOCOL

Phil Sayre opened the conference call by reviewing the field study protocol that the Subgroup has been developing. Sayre briefly outlined changes that he and Subgroup Co-Chair Lucinda Jackson have made to the protocol. A copy of the revised protocol has been included as an attachment to this document; copies will also be e-mailed to Subgroup members.

Species Selection

After reviewing the revised protocol, Sayre opened the discussion for comments. Steve Rock said that it would be interesting if the same fescue/ryegrass mix was used at every field site, thereby eliminating variations in plant species. Such a mix could act as a "benchmark mix," Rock said, though he acknowledged that choosing benchmark species for use at all sites would be tough.

Peter Kulakow suggested that it might be sufficient to specify only the variety of tall fescue and ryegrass, and not the species. Rock agreed. Jim Brown pointed out that difficulties could arise because of regional differences in varieties. While a case could be made for uniformity, he said, an equally good case could be made for choosing appropriate plant mixtures based on regional conditions. Rock replied that choosing native plants based on regional conditions is indeed one of the Subgroup's three treatment options, as outlined in the revised protocol:

    1) Tall fescue/perennial ryegrass mix
    2) Native grass mix or other species
    3) Unplanted control (to be kept weed-free)

Evelyn Drake then pointed out that it's only worth choosing a benchmark species mix for all sites if the species perform well; otherwise the field test becomes an experiment in how to make species grow in an inhospitable environment, which is obviously not the point. Rock suggested that the Subgroup could choose species that have been lab-tested and that grow everywhere. Drake replied that performance in the lab doesn't always parallel performance in the field.

Kulakow said that the tall fescue/ryegrass mix would probably be a good choice for quite a few locations; at other locations (e.g., Alaska), the logical choice would probably be red fescue—a substitute for tall fescue that is quite similar. He recommended using common sense in determining what will work at particular sites, while using comparable substitutes wherever possible.

Royal Nadeau pointed out the importance of specifying the precise volume of each species in the tall fescue/ryegrass mix. Rock suggested that the Subgroup could even add a third species to the mix. Drake said that researchers in Canada sometimes plant an annual rye to help establish cover in the first year while the perennial is taking hold; in subsequent years, the perennial rye/fescue mix is allowed to take over. Mike Reynolds said that an annual rye/red fescue mix was used at a site in Fairbanks, Alaska, and that the annual rye basically took over in the first year. (The site was reseeded the following year.) Kulakow added that there are some short-lived annuals that make good nurse crops, though he too has experienced situations where annual rye dominates a site.

Charlene Owens commented that the Subgroup might want to include a salt-tolerant species as a positive control in the benchmark mix. Participants agreed that the Subgroup could choose controls region by region, using one benchmark mix at one site and another mix at another site.

Kulakow stated that he sees the first treatment option—the fescue/ryegrass mix—as a "proved grasses" option, consisting of grasses that have some breeding history. He envisions the second treatment option as a "natural" treatment, consisting of native species or species that could be left in place at the site following treatment. He stated that the most important thing is to establish good cover, even if the same species can't be used at all sites.

Use of Clover/Alfalfa

Rock asked whether Subgroup members would approve of including alfalfa or clover in a species mix. Kulakow said that it makes a lot of sense; in some experiments at Kansas State University, alfalfa and clover have performed better than grasses. He recommended planting clover as part of a fescue/ryegrass mix, rather than planting it alone.

Drake pointed out that clover plantings have been damaged or wiped out by hydrocarbon at some sites. At a site in Canada, clover germination was affected by hydrocarbon concentrations of 10,000 parts per million. Kulakow said that he had seen similar effects. He added, however, that clover provides nitrogen fixation at sites where it thrives—another point in its favor.

Reynolds then asked whether the Subgroup should consider the succession of plants at field study sites. Ideally, he said, one would want to end up with a native mix or some kind of weed, even if one begins by planting grasses that have breeding histories.

Rock asked whether the Subgroup wants to use four components in a benchmark mix: annual rye, clover, perennial rye, and fescue. He suggested that the specific species could be juggled, while the percentages remain more or less the same. (For example, red fescue or another comparable grass could replace tall fescue at some sites.) Rock said that it would be interesting to see which species thrive where—which sites end up with fescue cover, which end up with clover cover, and so on. Another participant termed this treatment strategy "survival of the fittest."

Owens said that it might be worth running this idea by a botanist to check for potential problems. Kulakow commented that the Subgroup should be careful in determining the proportion of the ryes, as they tend to take over a site. He then suggested that the Subgroup choose between annual and perennial ryes at each site, since there probably wouldn't be many sites where one would want to plant both. He also recommended that instead of specifying clover as an ingredient in the benchmark mix, the Subgroup include the larger category of legumes. Many clovers are shallow-rooted, he said, and there may be sites where alfalfa (a legume) would be preferable.

Study Design Issues

In response to the discussion of developing a benchmark species mix, Drake pointed out that the Petroleum Environmental Research Forum (PERF) has designed its field study to include only a single species. She said that PERF could choose to plant two species, though this would cost more money and would mean that fewer sites could be investigated. She added that it would be in the best interest of the overall group—PERF and RTDF—if all sites could collect comparable data, though this might require EPA or someone else to help with analytical costs. Rock replied that he has been searching for funding sources but hasn't yet identified any. As the Subgroup progresses through the planning stages, Rock said, it will become easier to push through successful proposals.

Reynolds suggested that the Subgroup might want to arrange the criteria in the field study protocol into tiered groups, ranging from the essential to the not-so-important. He said that it might be valuable to establish the minimum criteria that each site should try to meet.

Drake agreed, saying that while it is useful to discuss the ideal design for the field studies, it is equally important to recognize that choices may have to be made. She added that, in her opinion, good root development and well-established cover are the priorities. In choosing between a native grass mix and the benchmark mix, one should consider which mix is more likely to achieve these goals.

Owens asked whether degradation of TPH, facilitated by plants, is the Subgroup's ultimate objective at each field site. She also asked whether any literature exists showing that perennial ryegrass and tall fescue aid in the degradation process. Drake responded that, in her experiments with these two species, the contaminant removal achieved through phytoremediation has been comparable to the removal achieved through other techniques (e.g., extensive tilling, in situ bioremediation). Her conclusion has been that the plants' root activity can take the place of extensive tilling. Owens then asked what the mechanism is that causes contaminant degradation; she pointed out that, in the early days of phytoremediation, it was thought that the nitrogen fixing caused by plants promoted increased microbial activity, which in turn resulted in degradation. Drake replied that researchers haven't seen any special effects caused by elevated microbial populations, relative to the effects that can be achieved through techniques such as extensive bioremediation. However, she said, some researchers now believe that one of the chief benefits of phytoremediation is that plants can help stabilize and elevate microbial populations over the long term, even after the intensive microbial activity and hydrocarbon removal cease. Kulakow verified that, at some sites, it can take as much as 2 years for the plants to establish their superiority over unvegetated control plots. Drake and Owens then agreed that it is important to include investigations of root density in the field study protocol. Drake said that it would be a "beautiful experiment" to include a comparison of root density at 6 months and 18 months.

Analysis Costs

Sayre commented that cost considerations seem to be an important factor for many of the Subgroup members. He asked whether the Subgroup will save money by arranging to have the samples from all sites analyzed simultaneously and by the same lab. He also suggested that it might be helpful to find out what the analysis costs might be.

Drake said that, from her discussions with Arthur D. Little (ADL), it sounded as though cost would be a function of the number of samples analyzed and the time period. According to Drake, ADL has the capacity to archive the total number of samples generated at all sites. (Drake estimated that the field study could generate at many as 48 samples per site over 3 years.) The charge per sample could be between $600 and $800 if the Subgroup chooses to analyze for TPH (saturated and aromatic hydrocarbon fractions), PAHs, and biomarkers. Drake offered to call ADL for more detailed pricing information.

Sayre stated that the Subgroup should consider how many sites it can reasonably afford to investigate, based on the number of samples to be generated and the likely analysis costs per sample. Kulakow said that the labs at Kansas State University have the capability to do TPH and basic soil analyses (though not PAH and biomarker analyses). The labs can complete analyses inexpensively and quickly, and they can handle numerous samples. Kulakow said that Kansas State could easily participate in the lab work, though the costs would obviously rise if all samples have to be sent out for PAH and biomarker analyses. Another participant pointed out that none of these cost estimates cover microbial or soil salinity analyses.

Rock said that it may be possible to split the analyses between two labs. Another participant asked whether the Subgroup could contract a university lab to do most of the analyses and ADL to do the more technical work. Drake then described a scenario in which each composite sample would be split, with part of the sample sent to a university lab for reasonably priced TPH analysis (perhaps $20 to $30 per sample) and part sent to ADL for archiving. Under these circumstances, Drake said, the Subgroup might be able to write a successful proposal specifically covering the analysis at ADL. Another participant approved of this idea, stating that it's not worth completing all of the expensive analyses if there's no indication of significant differences. Drake then commented that the Subgroup could get "the most bang for the buck" if, before undertaking the analyses, it identified which sites performed best.

Preparation of Samples

Reynolds asked whether soil samples would be split before being extracted and how this would be done. Another participant asked whether the samples would be dry or wet when split. Rock commented that the Subgroup should define these protocols carefully.

Drake said that, at her sites, staff usually sieve each sample to about 5 millimeters before doing TPH analysis. The soil is still moist during sieving, and it isn't actually ground up. The soil is sieved, mixed, then split.

Microbial Analysis

Drake asked the participants whether they want to specify a technique (such as heterotrophic bacteria most probable numbers [MPN]) for microbial analysis. Sayre said that Kathy Banks has described a technique to him in which she uses a redox dye to determine whether microbial respiration is going on. She also adds hydrocarbons directly as a carbon source. Instead of targeting heterotrophs in general, this technique specifically targets hydrocarbon-degrading bacteria. Sayre said he thought that one of Banks's graduate students had been working with Exxon on this technique. He offered to find out more information about Banks's technique and any associated costs.

Drake recommended using the same lab for all microbial analyses, so that the Subgroup could make good cross-sample comparisons. Reynolds pointed out that some Subgroup members might be more likely to do lab work "in house" or at a local university, while others might want to send samples out for analyses.

Reynolds said that the U.S. Army Cold Regions group plans to concentrate intensively on microbial issues and has drafted a protocol for an MPN method for hydrocarbon degraders. Reynolds commented that it would be useful to try to get some initial data showing that populations of hydrocarbon degraders tend to increase in the presence of plants. He said that the Cold Regions group has developed some preliminary data of this nature.

Drake recommended developing a standard protocol that could be used at a variety of labs or at some central location, such as Kansas State University. Kulakow said that he knew the name of the graduate student at Kansas State who had worked on the MPN technique. The student is currently finishing up her work at the university, but may be able to help coordinate microbial analyses in the future.

Owens suggested that those people with microbial experience work together to develop a detailed recommendation of the best possible analysis technique. Drake recommended that the Subgroup schedule a separate conference call specifically for discussing microbial analysis. Sayre said that he would try to arrange such a call. The participants agreed that Drake, Sayre, and Reynolds would participate in the microbial analysis discussion, along with Banks and Duane Wolf, if available.


SITE SELECTION

The conference call participants went on to discuss the criteria for selecting field study sites. Rock recommended that all sites have similar contamination—refinery waste is the major constituent among the sites proposed so far—and that the Subgroup shoot for a reasonable geographic spread. Beyond these criteria, Rock said, it's simply a question of how many sites the Subgroup can handle, both financially and organizationally. Sayre added that the Subgroup may want to consider selecting some sites with special characteristics—for example, high salinity, cold climates, or proximity to wetlands.

Rock then asked how many sites the Subgroup might want to include. Reynolds replied that that decision would depend on funding. Right now, he said, most participants are taking care of their own funding for their own sites. If Rock could identify some funds for analytical work, the Subgroup might be able to include another selection of sites.

Sayre asked whether the Subgroup wanted to include at least two sites that are fairly similar in terms of rainfall, climate, and other characteristics. Drake pointed out that the majority of the proposed sites are located within the same few regions where most participants operate—i.e., the Gulf Coast, California, and the East Coast. Sayre suggested that if the Subgroup included two sites where the benchmark grasses would grow similarly and provide similar results, it might be able to develop a more convincing argument. Drake said that her attitude is "the more sites, the better."


DATES FOR UPCOMING CONFERENCE CALLS

The next conference call is tentatively scheduled for Friday, April 24. A special conference call for issues pertaining to microbial analysis is being scheduled for Monday, April 20, at 11:00 or 11:30 a.m. (Eastern Time).


ACTION ITEMS


Attachment A

Field Study Protocol


REMEDIATION TECHNOLOGIES DEVELOPMENT FORUM
PHYTOREMEDIATION ACTION TEAM
FIELD STUDY PROTOCOL


TITLE: PHYTOREMEDIATION OF PETROLEUM HYDROCARBONS IN SOIL

PURPOSE: Determine efficacy of vegetating with agricultural and native grasses for degradation of petroleum hydrocarbons in soil at multiple locations and climatic conditions

TREATMENTS:

  1. Mixture of:

    1. Annual rye;
    2. A legume;
    3. Perennial rye; and
    4. Fescue.
      Proportions of each of four constituents adjusted to site.
      (??? salt-tolerant control??)

  2. Native grass mix or other species, including trees

  3. Unplanted control (kept weed-free)

Note: If funds are limited, treatment 1 or 2 could be dropped to favor the treatment that provides the best ground cover and dense root zone. Some concern is noted for the allelopathic effects of rye on other plants in the item 1 mixture. Legumes discussed for the mixture included alfalfa, clover, and bird's-foot trefoil.

LOCATIONS: To be determined.

SOIL SAMPLING: Take soil samples in each location at the following times:
  1. 0: Before sampling (samples taken at 0 to 6" and 1 to 2')
  2. 1: 6 months after planting (samples taken at 0 to 6" only)
  3. 2: 18 months after planting (samples taken at 0 to 6" only)
  4. 3: 30 months after planting (samples taken at 0 to 6" and 1 to 2')

The soil sampling and storage methods should meet specified standards. A 1" diameter GEOPROBE "large bore" sampler is suggested. Place each sample in a clean plastic sample liner.

PLANT SAMPLING: Take plant shoot samples only at time 3 (30 months) for hydrocarbon analysis.

SAMPLE STORAGE: Store samples at liquid nitrogen temperatures.

SAMPLE ANALYSIS: Send composite soil samples (after wet-sieving) to Arthur D. Little, Inc. (ADL), Environmental Monitoring and Analysis Unit, for archiving and analysis of:
  1. pH and salinity (times 0 and 3)
  2. Available nutrients (times 0 and 3)
  3. Petroleum (EPA method for saturated and aromatic hydrocarbons analyzed with GC, TPH measured, individual PAHs measured)
    (times 0, 1, 2, and 3)
  4. Biomarkers (times 0, 1, 2, and 3)
  5. Microbial analysis (times 0 and 3)
  6. Soil analysis (texture, organic matter, EC, CEC, soil type, etc.) (time 0)
Note: In tiered approach to sample analysis, samples would be archived at ADL. TPH would be measured and, if changes were insignificant (at 3-year point or earlier), further analyses as noted in items 1 to 5 would not be done. For sites that show significant TPH reductions in soil, samples could be selected from the archives for full analysis. TPH analyses could be done by ADL, or other labs using the ADL protocol.

PLOT SIZE: 20' x 20' minimum

REPLICATIONS: 4

STATISTICAL DESIGN: Randomized complete block based on presampling. For soil and plant samples, take four random sample cores per plot and make a composite sample.

GROWING CONDITIONS: Conduct site-specific seed bed preparation, planting technique, planting rate, and irrigation to establish good stand growth. Use record procedures.

Fertilization based on need from initial soil sampling.

PLANT EVALUATIONS: Evaluate plant characteristics at 6 months, 18 months, and 30 months:

  • Percent cover
  • Shoot height
  • Rooting characteristics (root depth and density)

WEATHER CONDITIONS:

Record rainfall throughout growing season.

SITE HISTORY: Source of contamination
Past site uses
Previous cleanup attempts
Climate