frequency, analyses, and interpretation are site specific and are key factors in accepting this
technology. Understanding the mechanisms and the limitations at a particular site are essential.
6.5 Lessons Learned
Key points are:
The selected plant needs to grow. The cleanup mechanism is a root-surface phenomenon.
Grasses, which have fine, dense roots with high surface-areas, are acceptable. Plants are most
susceptible to stress during the seedling stage. After establishment, volunteer plants typically
establish themselves. If the soil is too contaminated or over fertilized, simply waiting for
volatilization or microbial incorporation of the excess may be sufficient for conditions to become
conducive to seed germination and establishment.
Although there may be exceptionally good and exceptionally poor plants for enhancing
petroleum degradation, they have not all been identified. Extensive plant screening is difficult
and costly, and results probably vary with many other conditions such as temperature, the nature
of the petroleum, soil conditions, rainfall, and other conditions not yet understood or identified.
The University of Saskatchewan has developed a database, PhytoPet
(http://www.phytopet.usask.ca/mainpg.php), to catalogue plants for petroleum phytoremediation.
PhytoPet was originally developed as an inventory of plants that have demonstrated ability to
either phytoremediate or tolerate soils contaminated with petroleum hydrocarbons. As with much
phytoremediation information, the database is changing and allows for user interaction. There
also are molecular-based efforts that are attempting to screen plants by looking for specific genes
in plants and matching these to contaminant degradation pathways, but this research is not yet to
the application stage.
Petroleum degradation is well characterized, and for rhizosphere-enhanced remediation the
process is a root-surface phenomenon, rather than one centered in the plant. From CRREL's
experience, grasses do well for petroleum. This is most likely due to their fibrous root system
that explores a large volume of soil fairly completely and, in a sense, provides pseudo-mixing. In
various field studies at other sites, we also have used annual ryegrass (Lolium multiflorum), tall
fescue (Festuca arundinaceae), and winter rye (Secale cereale L.). We have seeded at rates
heavier than would be used normally for establishing the grass. Extra seed is to account for
losses from poor germination and seedling die-off due to petroleum contamination and poor
growth conditions, such as drought. The goal is to get a good plant cover on the soil and
thorough root growth and penetration in the soil.
Fertilization is important. It is
easy to
over fertilize using commonly cited carbon:nitrogen
ratios as a target. Maximal fertilization levels are more a function of soil texture and soil water
holding capacity than soil contamination levels. Recent data suggest that fertilization alone
(without plants) can inhibit depletion of some petroleum fractions.
There are proprietary fertilizers on the market, specifically aimed at bioremediation and
phytoremediation. Data supporting the benefits of these products are quite scarce and often not
critically defensible. For example, CRREL reviewed the marketing literature for a product
marketed as a "petroleum remediation enhancer" that showed graphs of concentrations
decreasing with time. However, the petroleum was jet fuel, the soil was sand, it was tilled every
day, it was hot and windy, and there were no control treatments for comparison. Most of the
28