The fern that started the field — Pteris vittata, the first arsenic hyperaccumulator ever described, with frond As above 2% dry weight in its original 2001 report — turns out to have a second, quite different metal career. Wan, Qian and colleagues (Guizhou University / Institute of Geochemistry, Chinese Academy of Sciences) report in the Journal of Hazardous Materials (517:143479, published online 2026-09-02) a field-and-greenhouse study of the fern in abandoned mercury mining areas. The field numbers first: plants growing on those mine sites reached total mercury concentrations of up to 163 mg/kg in roots and 83 mg/kg in shoots. Those are “up to” site maxima, not means — but they show the fern takes up substantial Hg from real, contaminated ground, not just from spiked hydroponics.
The mechanism is sulfide, not the arsenic playbook
Here is the non-obvious part. P. vittata’s fame rests on arsenic: it reduces arsenate to arsenite, binds it to thiols and pumps the complexes into vacuoles via the ACR3 transporter — a story of active compartmentalisation. X-ray absorption near-edge structure (XANES) analysis in this study says mercury is handled completely differently: the predominant Hg species in both roots and shoots were β-HgS (metacinnabar, the black sulfide) and Hg(SG)₂, the bis-glutathionato complex — the authors’ own interpretation is “inorganic precipitation and thiol-mediated complexation as key detoxification pathways”. Subcellular fractionation put most Hg in the operationally defined soluble fraction and the cell walls of both leaves and roots.
One chemical caveat belongs here, because a careless reading gets the mechanism wrong. β-HgS is among the most insoluble compounds known — it cannot be literally dissolved in cell sap. XANES identifies mercury’s bonding, not its address: the sulfide detected in the “soluble” fraction most plausibly reflects colloidal- or nanoparticulate-scale sulfide that survives centrifugation, or sulfide formed or redistributed during tissue homogenisation. The defensible reading of the two results together is sequestration as extremely fine solid sulfide plus glutathione complexation — mercury chemically locked down, not mercury dissolved and managed.
The failure mode: photosynthesis falls first
The stress-physiology half of the paper carries its own warning. Under HgCl₂ stress, glutathione and malondialdehyde contents increased — the classic antioxidant-mobilisation and lipid-peroxidation signature — while peroxidase activity and proline first rose and then declined as exposure continued, the signature of a defence being overwhelmed. The ordering result is the interesting one: chlorophyll decline began earlier than the proline and peroxidase reductions, meaning the photosynthetic system is the first domino, not the last. At the extreme dose of 1000 mg/kg HgCl₂, transmission electron microscopy showed severe cellular and ultrastructural damage. For field biomonitoring that ordering is a usable early-warning signal — chlorophyll as the cheapest Hg-stress indicator this fern offers — but the dose that produced the damage is orders of magnitude above bioavailable Hg in nearly any real soil, and it is a soluble HgCl₂ spike, not the sulfide-dominated speciation the authors measured in the field. The greenhouse stress results bound the mechanism; they do not describe it.
What it does not show
The paper’s hedged label — “potential mercury hyperaccumulator” — is doing honest work, and a reader should notice three things it quietly concedes. First, there is no canonical hyperaccumulation threshold for mercury: the Baker & Brooks framework and its modern successors assign cutoffs for Cd, Tl, Se, Co, Cu, Cr, Ni, As, Pb, Zn and Mn, but not Hg; a co-author’s own 2018 review stated that no recognised mercury hyperaccumulator had been identified and the threshold remained uncertain. Against the ~100 mg/kg operational criterion some of the Chinese Hg literature uses, the 83 mg/kg shoot maximum here would not qualify. Second, roots out-concentrate shoots (163 vs 83 mg/kg), so the translocation factor is below one: classic phytoextraction harvests above-ground biomass, and most of the mercury in this fern stays below ground, in tissue that would have to be dug up and handled as hazardous waste. Third, this is not a first: US–Chinese groups screened P. vittata for mercury phytoextraction as early as 2008, and hydroponic uptake studies followed in 2009; Wan et al.'s contribution is mechanistic depth — field validation in abandoned mines, subcellular fractionation, XANES speciation — not discovery. Nor does the abstract address the two confounds that haunt mercury phytoremediation claims: foliar uptake of gaseous Hg(0) from air and foliar re-emission, either of which can inflate or misattribute shoot Hg measured in the field.
The honest use case is stabilisation, not harvest
Read the mechanism against the risk ledger and the piece writes itself. A plant that precipitates mercury as sulfide and walls it up in root tissue is not a phytoextractor; it is a phytostabiliser — and for mercury, where the bioavailable form that matters is methylmercury produced by soil microbes, binding Hg into sulfide is precisely the chemistry that keeps it out of the methylation pathway. That is the real opportunity in this paper: root-zone Hg converted to β-HgS and thiol complexes is Hg less available to methylators, anchored in a root system that can later be removed. The genuine risks are agronomic and evidentiary — the shoot numbers are site maxima from soils we know nothing about (no soil concentrations, BCFs or translocation factors appear in the abstract), the greenhouse stress doses are unrealistically high, and volatilisation confounds mean field shoot Hg is a noisy proxy for uptake. If mercury phytoremediation with this fern ever matures, the harvest will likely be roots, not fronds — a logistics problem the arsenic success story, which conveniently concentrates its metal in the aerial part, never had to solve. The deeper lesson generalises: the machinery that makes P. vittata a champion for arsenic — vacuolar thiol sequestration in fronds — is not what it uses for mercury, so cross-metal hope built on this species’ fame should be checked chemistry-first, species by species.
Source: Wan, Qian, Qiu, Ao, Wang, Liang, Han, Habibullah-Al-Mamun, Li, Wu, He & Xu (2026), Journal of Hazardous Materials 517:143479, DOI 10.1016/j.jhazmat.2026.143479 (published online 2026-09-02; accepted 2026-08-31). The article is closed access; this piece is therefore bounded to the peer-reviewed abstract, and every claim above is checkable against it. Every load-bearing number — the 163 mg/kg root and 83 mg/kg shoot maxima, the 1000 mg/kg HgCl₂ dose, the XANES speciation (β-HgS, Hg(SG)₂), the subcellular distribution, and the biomarker ordering (chlorophyll declining before proline/peroxidase) — was verified verbatim by three independent reviewers retrieving the Europe PMC/PubMed, Crossref, OpenAlex and Semantic Scholar records separately; no numeric discrepancies were found. The reviewers also flagged, and this text avoids, the abstract’s own singular-“shoot” typo; the β-HgS/soluble-fraction tension and the absence of a canonical Hg hyperaccumulation threshold are reviewer caveats reflected above. See /methodology/ for how we source and check analysis pieces.