An excluder on a calamine heap, in numbers
A new study in Antioxidants compares two ecotypes of Lotus corniculatus — bird’s-foot trefoil — a metallicolous population from a century-old zinc–lead “calamine” tailing at Bolesław, in the Olkusz mining district of southern Poland, and a non-metallicolous population from a limestone stonepit some 290 km away (Sujkowska-Rybkowska et al. 2026). The calamine substrate is extreme: the paper reports total soil Zn ≈ 43,000 mg kg⁻¹ (printed “42.983” in the European decimal convention), Pb ≈ 2,298 mg kg⁻¹ and Cd ≈ 184 mg kg⁻¹, against 55.6, 10.4 and 1.0 mg kg⁻¹ in the control soil.
Yet the calamine plants are not hyperaccumulators. All three metals show a translocation factor below 1 (shoot:root Zn 0.3, Pb 0.2, Cd 0.2), and shoot concentrations sit one to two orders of magnitude under hyperaccumulation thresholds: shoot Zn 344 ± 151, Pb 13.9 ± 3.1, Cd 2.8 ± 1.8 mg kg⁻¹ dry matter, with roughly three to five times more in the roots (root Zn 1,042, Pb 71, Cd 14 mg kg⁻¹). This is textbook metal exclusion: keep the burden in the roots, protect the shoot.
The mechanism: exclusion plus a rerouted redox economy
Exclusion is only half the story; the other half is how the shoot that still receives some metal stays functional. Relative to the control ecotype, the calamine plants ran less oxidative stress, not more — lower H₂O₂ and lower lipid peroxidation (MDA) — while restructuring their antioxidant system. Superoxide dismutase and catalase activities were about twice as high and glutathione peroxidase was elevated, but the classic ascorbate–glutathione recycling arm was throttled down: ascorbate peroxidase 3.3× lower, monodehydroascorbate reductase 6.5× lower, glutathione reductase 2.4× lower. At the same time the standing pool of reduced glutathione was roughly 2× higher (with less oxidised GSSG), and total phenolics rose — led by catechin, rutin (quercetin-3-O-rutinoside), and salicylic and ferulic acids, compounds that chelate metals as well as scavenge radicals.
The cleanest reading is not “the plant scavenges reactive oxygen harder” but that it has re-routed its redox economy toward a chelation-first strategy: hold a large reduced-glutathione pool, lean on phenolic chelators and on catalase/GPX for peroxide, and let the ascorbate–glutathione recycling loop idle because upstream sequestration keeps radical production low. A telling side-effect: the calamine chloroplasts held no starch grains, consistent with carbon and energy being diverted into detoxification, symbiosis and repair rather than storage — the metabolic rent of tolerance. (The authors flag one unresolved puzzle of their own: glutathione up while both GSSG and glutathione reductase are down implies extra de-novo synthesis or GSSG export, which they say needs further study.)
For phytoremediation, a stabiliser — not an extractor
An excluder is the wrong tool for phytoextraction and the right tool for phytostabilisation. Where a zinc/cadmium hyperaccumulator sharing this habitat type — Arabidopsis halleri, which grows on the same Bolesław–Bukowno wastes, or Noccaea praecox and N. caerulescens — pulls metal up into harvestable shoots for removal, a tolerant, nitrogen-fixing legume like L. corniculatus does the opposite job well: it can establish a self-sustaining, metal-poor canopy that armours the tailing surface and builds soil nitrogen without concentrating metal in its biomass. Independent field data support the excluder label — Hasnaoui et al. 2020 report translocation factors of about 0.55–0.60 for Cd, Pb and Zn in L. corniculatus on Moroccan Pb–Zn tailings. (Our earlier piece on kidney vetch, another calamine legume from the same research group’s system, found the mirror-image split: thallium hyperaccumulation but zinc exclusion.)
Two honest bounds. This study measured physiology, not remediation performance — no biomass, ground-cover or erosion data — so the species is suited to stabilisation, not shown to deliver it. And exclusion is not zero transfer: shoot metal still returns to the surface each year in litter, a stabilisation-versus-dispersal trade-off worth watching.
The catch: a forage legume over the cadmium feed limit
Here is the non-obvious risk, and it is ours rather than the authors’. Lotus corniculatus is not an obscure metallophyte; it is bird’s-foot trefoil, one of the most widely sown pasture and hay legumes of the temperate world. A calamine ecotype — or a cultivar bred to carry its tolerance — deployed on contaminated land would look healthy and green (the paper reports “no toxic symptoms at the ultrastructural level”) while still moving cadmium into the food chain. At 2.8 mg kg⁻¹ dry weight, shoot Cd is about 2.5 mg kg⁻¹ on the 12%-moisture basis EU feed law uses — roughly 2.5× the 1 mg kg⁻¹ maximum for feed materials of plant origin under Directive 2002/32/EC, which grants no forage exemption for cadmium. Lead is the reassuring contrast: shoot Pb of about 12 mg kg⁻¹ at 12% moisture exceeds the generic 10 mg kg⁻¹ feed-material limit but stays well under the 30 mg kg⁻¹ limit for forage, so the clean regulatory flag is cadmium, not lead.
The mechanism sharpens the hazard rather than softening it. The paper’s Evans-blue staining shows dead cells concentrated in the oldest leaves — metal dumped into senescing tissue, the disposal trick seen in other calamine plants — and those old leaves stay attached and grazeable. “Tolerant and green” is not the same as “safe to graze.” Two caveats keep this fair: the risk is sharpest for a sown cultivar on moderately contaminated ground, because on a bare tailing (soil Cd here is 184 mg kg⁻¹) direct soil ingestion usually dominates livestock exposure anyway; and the shoot-Cd mean carries a wide spread (2.8 ± 1.8 mg kg⁻¹, n = 3), so individual plants likely range from around the limit to several times over it.
How far to trust it
This is a two-population field comparison, and the sites differ in substrate, climate and microbiome as well as genotype, so “adaptation” is inferred, not proven — there was no common garden or reciprocal transplant to separate heritable differentiation from lifelong acclimation, and several assays rest on n = 3 (phenolics and antioxidant-capacity assays used n ≥ 6). The soil totals were not measured in this study but carried over from the group’s earlier work, and the calamine figures do not match the specific paper cited for them (another of the group’s papers reports >10,000 mg kg⁻¹ Zn, 1,561 Pb and 129 Cd for the same site), a reminder of how spatially heterogeneous tailings are; treat the soil numbers as order-of-magnitude. Exclusion in this species is typical but not universal — at least one study from another mine district claims the opposite, most likely a dust-contamination artifact. What is solid is the shoot-tissue chemistry, and it carries two implications at once: a promising phytostabilisation candidate, and a cadmium-carrying forage plant that would pass a visual health check.
Provenance: based on the full open-access text (CC BY) of Sujkowska-Rybkowska et al. (2026), Antioxidants 15(8):969. Every tissue, soil, enzyme and antioxidant figure was checked verbatim against the paper’s Table 1 and Results text by three independent reviewers, the metadata against Crossref and OpenAlex, and the EU limits against Directive 2002/32/EC. See our methodology for how we verify.