Quality control of pore water chemical analyses is intrinsically difficult.
To the best of authors' knowledge, there are no specific certified international
standards, and replicate analyses were seldom possible.
Preliminary release tests with deionised water on new lysimeters and
rhizometers indicated negligible (below detection limits for ICP–OES)
amounts of released elements, except for Ba. For this element, high
values were recorded in lysimeter blanks, therefore data for Ba were
not included in this study. Field blanks were obtained by introducing
deionised water in the same devices used for pore water samples, and
treating the collected volumes as independent samples. The accuracy
of ICP–OES analyses was checked with calibration standard solutions.
A further check of quality of analysis for major elements is ionic balances.
From data in Table 3b, it is possible to calculate ionic balances:
most are within ±10%, and only two are within ±15%. Overall, it is estimated
that most of analyses are within ±20% error for both accuracy
and precision.
The collected data were elaborated with two different speciation
programs, PHREEQC 3.0.2 (Parkhurst, 1995; Parkhurst et al., 1980)
and WHAM 7.0.2 (Tipping, 1994); specifically, the second accounts for
complexation of metals by organic matter.