3. Demonstration Design
3.1 Performance Objectives
The objective of this effort was to demonstrate rhizosphere-enhanced bioremediation of
petroleum-contaminated soils located in cold, remote sites. We measured success by examining
changes in the composition as well as concentration of petroleum in the soils.
Due to variability inherent in field data and the relatively slow treatment rates in cold regions,
obtaining sufficiently precise field data to measure treatment effects on contaminant
concentration is exceedingly difficult. Those involved in petroleum phytoremediation generally
agree that the primary mechanism for phytoremediation of petroleum compounds is increased
microbial activity in the rhizosphere rather than plant uptake, as is often erroneously assumed.
As described in Section 2.3, our laboratory and field studies suggest that the rhizosphere effect is
increasingly important as the recalcitrance of the compound in question increases (Reynolds et
al., 1999; Reynolds et al., 1997). Although the enhancement due to a rhizosphere effect, relative
to non-vegetated soil, is likely greatest for heavier, more recalcitrant compounds, the resistance
to degradation of these heavier compounds may result in longer treatment times being required
before rhizosphere effects can be measured.
One approach is to monitor petroleum concentration changes in each treatment. At present, the
final measure of performance is reduction of contaminant concentrations in the soil. We did not
expect to attain concentrations that were asymptotic to a field endpoint at the end of this
demonstration. To help address this, we used biomarker techniques to evaluate changes in the
In brief, this
approach compares relatively degradable fractions
petroleum to those that are recalcitrant. Highly weathered petroleum will have a high percentage
of recalcitrant compounds compared to fresh or moderately weathered petroleum product. We
monitored changes in fraction specific hydrocarbons (FSH)--an approach that attempts to
classify hydrocarbons by grouping them into functionally similar fractions. Because of their
functional similarity, the fractions can be separated by extraction and clean-up procedures. The
fractions were also delineated so that there is toxicity data on at least one compound in each
fraction. The assumption is that the toxicities of compounds within a fraction are more similar
than across fractions, and therefore within-fraction toxicity data is the best estimate to use for
extrapolating to compounds lacking toxicity data.
Table 1 summarizes our performance objectives and how they were met.