A five-year, three-site answer to a scaling question
Pfendler, Ciadamidaro, Ozaki, Zappelini, Blaudez and Chalot (2026), at the Université Marie et Louis Pasteur and CNRS in Montbéliard, France, ran one of the more ambitious woody phytomanagement experiments in the literature: 31 woody species established at three French trace-element-contaminated sites — Thann, Carrières-sous-Poissy and Leforest — and followed for five years (doi:10.1016/j.ecoenv.2026.120820, Ecotoxicology and Environmental Safety 324:120820, published online 29 September 2026). The sites differ in soil properties and contamination history; above- and belowground fresh biomasses were quantified, rhizosphere pH and CaCl₂-extractable elements were measured, and trace elements were analysed in leaves, bark and trunk of the two species common to all three sites, Salix aquatica ‘Grandis’ and Ostrya carpinifolia.
The winner rotates with the site
The headline biomass result is instability, not a champion. At Carrières-sous-Poissy, Elaeagnus umbellata and Acer saccharinum produced the highest fresh biomass; at Leforest, Salix alba and Alnus subcordata performed best; at Thann, Ostrya carpinifolia and Salix alba topped the ranking. Only S. alba appears on two podia, and no species led everywhere. Shoot and root fresh biomasses were strongly correlated at each site — a useful operational result, since it means belowground performance can be inferred without excavating roots. The authors’ own conclusion is blunt: woody-species selection for contaminated land must be site-specific.
High cadmium and zinc — and still below the line
The chemistry sharpens the agromining question. Rhizosphere soil pH showed only non-significant numerical trends among species, while selected CaCl₂-extractable elements differed between planted groups and site-specific controls — so the planting itself is measurably perturbing the bioavailable pool even where pH is silent. The two sentinel species split the roles. Ostrya carpinifolia carried comparatively low trace-element concentrations in all tissues — a phytostabilizer and biomonitor profile, useful for containing the contaminant load while signalling site chemistry. Salix aquatica ‘Grandis’ went the other way: it accumulated high Cd and Zn concentrations in leaves and bark — the abstract’s qualitative “high”, with no tissue values printed — yet its maximum mean foliar concentrations still remained below the operational hyperaccumulation thresholds for both elements. In the authors’ framing these are complementary candidates for phytoextraction, biomonitoring and phytostabilization, not a single all-purpose tree.
The point is easy to misread, so it is worth stating carefully. This is not “willows fail at metal removal”; it is that a very high-biomass, moderately concentrating tree clears soil by a different arithmetic than a textbook hyperaccumulator. Our own thresholds, following van der Ent and colleagues, sit at 100 µg g⁻¹ for cadmium and 3,000 µg g⁻¹ for zinc — the benchmark the paper’s “operational” thresholds are in dialogue with, though the abstract does not print its cut-off values. Contrast the true hyperaccumulators already in the database — Arabidopsis halleri or Sedum alfredii, which reach percent-level Zn in shoots at a fraction of the biomass. Phytoremediation economics is biomass × concentration × harvests per year; this trial says that on contaminated French brownfield soils, the first factor is where woody systems win, and that a sub-threshold leaf can be a price worth paying for a very large harvestable biomass — provided the wood itself is then handled as a metal-bearing product, not fuel.
Opportunity and risk
The opportunity is a portfolio instead of a silver bullet: biomass species chosen per site, a sentinel species for monitoring, and a stabilizer for the edges — all from one five-year dataset spanning genuinely different soils. The risks are equally concrete. Sub-threshold foliage means slow soil decontamination: willow phytoextraction is a decades-scale commitment, and anyone selling it as rapid remediation is overselling. A tree that concentrates Cd and Zn in bark and leaves produces a contaminated harvest stream whose valorization (or safe disposal) is a cost line, not an afterthought — the moment the wood is burned or mulched, the metals are back in circulation. And a site-specific winner is also a site-specific unknown: transplanting the Thann or Leforest rankings to a fourth site is exactly the extrapolation this paper argues against.
Bottom line
Five years and three contaminated sites say that site beats species in woody phytomanagement, that the most conspicuous Cd–Zn accumulator on trial (Salix aquatica ‘Grandis’) does its work through biomass rather than hyperaccumulation, and that honest contaminated-land forestry is a complement to — not a substitute for — the true hyperaccumulators that anchor this database.
Provenance: this paper is subscription-access; the analysis is bounded to the complete PubMed/Europe PMC abstract (PMID 42810263) and Crossref metadata, which contain no absolute tissue concentrations — accordingly, no tissue values are quoted. The paper’s title is quoted exactly (“phytoremediation potential … field trials”). Every load-bearing claim was verified verbatim against the primary-source abstract by three independent reviewers on separate vendors (Claude / GPT / Kimi, each self-retrieving from Europe PMC and Crossref); all 20 checked claims were confirmed by all three, with one title-wording correction caught and fixed before publication. See /methodology/.