The main purpose of this study was to determine typical concentrations of heavy metals (HM) in wood from willows and poplars, in order to test the feasibility of phytoscreening and phytoextraction of HM. Samples were taken from one strongly, one moderately, and one slightly polluted site and from three reference sites. Wood from both tree species had similar background concentrations at 0.5 mg kg−1 for cadmium (Cd), 1.6 mg kg−1 for copper (Cu), 0.3 mg kg−1 for nickel (Ni), and 25 mg kg−1 for zinc (Zn). Concentrations of chromium (Cr) and lead (Pb) were below or close to detection limit. Concentrations in wood from the highly polluted site were significantly elevated, compared to references, in particular for willow. The conclusion from these results is that tree coring could be used successfully to identify strongly heavy metal-polluted soil for Cd, Cu, Ni, Zn, and that willow trees were superior to poplars, except when screening for Ni. Phytoextraction of HMs was quantified from measured concentration in wood at the most polluted site. Extraction efficiencies were best for willows and Cd, but below 0.5 % over 10 years, and below 1 ‰ in 10 years for all other HMs.Electronic supplementary materialThe online version of this article (doi:10.1007/s11356-013-2085-z) contains supplementary material, which is available to authorized users.

This paper illustrates the usefulness of pre-screening methods for an effective characterization of polluted sites. We applied a sequence of site characterization methods to a former Soviet military airbase with likely fuel and benzene, toluene, ethylbenzene, and xylene (BTEX) contamination in shallow groundwater and subsoil. The methods were (i) phytoscreening with tree cores; (ii) soil gas measurements for CH4, O2, and photoionization detector (PID); (iii) direct-push with membrane interface probe (MIP) and laser-induced fluorescence (LIF) sensors; (iv) direct-push sampling; and (v) sampling from soil and from groundwater monitoring wells. Phytoscreening and soil gas measurements are rapid and inexpensive pre-screening methods. Both indicated subsurface pollution and hot spots successfully. The direct-push sensors yielded 3D information about the extension and the volume of the subsurface plume. This study also expanded the applicability of tree coring to BTEX compounds and tested the use of high-resolution direct-push sensors for light hydrocarbons. Comparison of screening results to results from conventional soil and groundwater sampling yielded in most cases high rank correlation and confirmed the findings. The large-scale application of non- or low-invasive pre-screening can be of help in directing and focusing the subsequent, more expensive investigation methods. The rapid pre-screening methods also yielded useful information about potential remediation methods. Overall, we see several benefits of a stepwise screening and site characterization scheme, which we propose in conclusion.Electronic supplementary materialThe online version of this article (doi:10.1007/s11356-015-4649-6) contains supplementary material, which is available to authorized users.

Abstract. 19Tree core samples have been used to delineate organic subsurface plumes. In 2009 and 2010, 20 samples were taken from trees growing on a former dump site in Norway and analyzed for 21 arsenic(As), cadmium(Cd), chromium(Cr), copper(Cu), nickel(Ni) and zinc(Zn). 22Concentrations in soil were in averages 30 mg/kg dw for Zn, 2 mg/kg dw for Cu, and < 1 23 mg/kg dw for Cd, Cr, As and Ni. The concentrations in wood samples from the polluted test 24 site were compared to those derived from a reference site. For all except one case, mean 25 concentrations from the test site were higher than those from the reference site, but the 26 difference was small and not always significant. Differences between tree species were 27 usually higher than differences between reference and test site. Furthermore, all these 28 elements occur naturally, and Cu, Ni and Zn are essential minerals. Thus, all trees will have a 29 natural background of these elements, and the occurrence alone does not indicate soil 30 pollution. For the interpretation of the results, a comparison to wood samples from an 31 unpolluted reference site with same species and similar soil conditions is required. This 32 makes the tree core screening method less reliable for heavy metals than, e.g., for chlorinated 33 solvents. 34 35 36 37

Preliminary risk assessment for prioritisation of site investigations requires efficient screening to reveal type and level of contamination. The screening methods, tree coring and soil gas sampling were applied and compared at two forested sites contaminated with tetrachloroethylene (PCE) or trichloroethylene (TCE) to evaluate their ability to locate source zones and contaminant hot spots. One test site represented a relatively homogeneous sandy soil and aquifer, and the second a more heterogeneous geology with both sandy and less permeable clay till layers overlying a chalk aquifer. Tree cores from different tree species were sampled and analysed, and compared to soil gas measurements and existing soil gas data. Both methods were found useful as screening tools to locate hot spots of PCE and TCE in the shallow subsurface. Tree coring was found to be particularly beneficial as a complement to soil gas sampling at sites with low permeable soils, and where contamination was located in the capillary rise or shallow groundwater. The shorter time required for tree coring reduced the costs compared to soil gas sampling, but the sensitivity and precision of tree coring were lower. However, this did not affect the feasibility of using tree coring to locate the hot spots. Moreover, a combination of the two methods can help to focus any subsequent investigations like soil or groundwater sampling. The use of tree coring to complement soil gas sampling for pre‐screening is expected to result in higher certainty for revealing hot spots and source zones at contaminated sites.