Page 41 - Plant-Based Remediation Processes

28 M. Barbafieri et al.

reduction in soil after a phytoremediation treatment. Pot experiments by Ye et al.
2011 showed a reduction of about 11–38 % after 9-month period of Pteris vittata
growing for arsenic potentially available (phosphate extractable) and 18–77 % in
soil pore water As. Tassi et al. 2011 reported a reduction of 45 % of bioavailable
boron after two consecutive growing cycles in microcosm test. Cassina et al. 2012
reported a reduction of 33–45 % of mobile mercury after one growing of H. annuus
and B. juncea respectively in microcosm pot simultaneously treated by cytokinin
and thiosulphate. In field experiments this approach is often not considered. The
main cause is the high heterogeneity of metal distribution in contaminated soil.
Blaylock and Elless 2009 reported a 5-year field study on arsenic removal. But after
different sampling grid conducted after each growing season to verify the arsenic
removal from soil, they do not observe a significant arsenic removal due to the high
soil heterogeneity in arsenic content. The sampling variability challenges the
phytoremediation evaluation when approaching a study of mass balance in field
experiments (Brus et al. 2009; Van Nevel et al. 2007).

2.2.4 Decision Support Systems

For phytoextraction, a very important critical success factor is the duration of a
phytoextraction, i.e., the period between starting the process and the moment when
the total concentration or the bioavailable concentration of heavy metal(s) has
reached regulatory target levels for soils (Koopmas et al. 2007). To use the total
or the bioavailable concentration as target value depends on the legislator’s
demands; total or bioavailable fractions are determined by standard extraction
procedures, e.g., a diluted calcium chloride extraction to mimic plant-availability
(Ro ¨mkens et al. 2009). Many authors simply use a linear phytoextraction model in
which the amount of phytoextracted heavy metal is assumed to be independent of
the actual heavy metal concentration in soil or soil solution at a certain stage during
phytoextraction. Such an approach is definitely a gross simplification which in most
cases will underestimate the real phytoextraction duration. It is more probable that
the phytoextraction rate in the case of non-hyperaccumulators depends on the actual
supply of plant-available heavy metals in the soil, which steadily decreases during
the phytoextraction duration. In the case of hyperaccumulators the story might be
different; as uptake by such plant species is assumed to be (not only) supply-driven,
as “active” processes in the plant root zone may play a role as well. Anyway it is not
very likely that a simple model can easily predict phytoextraction duration for both
types of plants. Instead of this an experimental protocol can be used, based on
mixing the polluted soil with different amounts of clean soil with the same general
composition and determine after a period of aging both the plant-available heavy
metal concentration (by chemical extraction) and the actual uptake by the chosen
phytoextraction plant species. Albeit time-consuming (several months), it results in
a better prediction of phytoextraction duration than just using a linear model.
Results of such tests also confirm the hypothesis that a nonlinear model is more