Odontarrhena is the genus non-specialists reach for when they want a nickel hyperaccumulator: of the ~532 Ni hyperaccumulators among the 721 metal-hyperaccumulating species catalogued by Reeves et al. (2018, New Phytologist), the Brassicaceae — and Odontarrhena in particular — hold the largest concentration (some four dozen species in older counts, by Brooks 1998). Our own database is built on that genus: O. chalcidica and O. serpyllifolia anchor the nickel hub. So a new paper testing an Odontarrhena that grows only on non-serpentine soils was always going to be interesting — and the answer it gives, in Pavlova, Nacheva, Karadjova, Tsonev and Bani (2026, Ecological Research 41(5)), is the one a good hypothesis predicts: the species does not hyperaccumulate. It excludes.
Three congeners, three substrates, one trait gap
The Bulgarian–Albanian team compared three species across a deliberately edaphic design. In the field they sampled O. orbelica — a Bulgarian endemic of the Pirin Mountains, listed as Critically Endangered in the Bulgarian Red Data Book under its basionym Alyssum orbelicum — alongside O. muralis, both on non-serpentine soils, plus O. decipiens on serpentine. Then came a pot experiment across three serpentine soils, one carbonate soil and one volcanic soil, and a hydroponic series of NiSO₄ at 10, 25, 50, 100, 500, 1000 and 3000 µM, with Ni (plus Fe, Mn, Zn, Ca and Mg) measured by ICP-OES in roots, stems and leaves and root-to-leaf translocation and bioaccumulation factors computed throughout. The results, per the publisher’s abstract: O. orbelica “behaves as a Ni excluder” with very low tolerance of elevated Ni; O. decipiens “exhibited characteristics of a hyperaccumulator in both the pot and hydroponic experiments”; and O. muralis “accumulated Ni in serpentine soils” while tolerating Ni up to what the abstract reports as approximately 2000 µM.
The negative result is the mechanism
Read against the elemental-defense account of hyperaccumulation — Ni at
1,000 µg g⁻¹ in leaf dry matter is toxic to most herbivores and pathogens, so the trait pays for itself only where soil Ni is high enough to make exclusion costly — an excluder on carbonate soils is not an anomaly but a prediction. Serpentine soils both supply the selective pressure and reward the investment; off-serpentine, the same physiology is dead weight. O. orbelica is not even the first such case: O. sibirica, another non-hyperaccumulating member of the genus, was documented as unable to accumulate Ni by Bettarini et al. (2020, Planta), who titled it a “paradox” and resolved it the same way — the trait tracks the substrate, not the clade. Two excluder species is not yet a genus-wide phylogeny, but the direction is consistent: hyperaccumulation in Odontarrhena looks edaphically coupled, acquired (or retained) where serpentine made it pay.
The agromining read: substrate history beats genus lists
For Ni agromining this is a selection warning, not a curiosity. O. muralis is the Balkan “metal crop” most often proposed for Ni phytomining, yet Bettarini et al. found non-serpentine populations of that same species unable to accumulate Ni — accumulation there is population- and substrate-dependent. Pavlova et al. add the complement: a congeneric endemic whose entire range is non-serpentine never had the trait. The practical rule for anyone building a candidate list from genus-level databases (including ours) is that provenance is part of the phenotype: a serpentine-collection history is evidence, a congeneric label is not. Screen on the soil you intend to mine, because the plant’s accumulation physiology was written by the soil it came from.
The conservation read runs the other way
Here is the less obvious implication. Serpentine endemics enjoy a “toxic-soil refuge”: their habitat is hostile to crops and most competitors, which shelters them from agriculture. O. orbelica has no such refuge. It lives on ordinary carbonate-derived mountain soils — land that can be converted, fertilized, built on — and it is Critically Endangered. If hyperaccumulation is edaphically coupled, then the Odontarrhena lineage’s conservation portfolio is not confined to serpentine barrens; some of its most threatened members live where the soils are benign and the land competition is real. Metallophyte conservation frameworks that only watch ultramafic sites would miss them.
What we could and could not verify
This paper is closed access, and the publisher full text is bot-gated even to readers, so we verified everything against the publisher-deposited abstract, retrieved independently through three routes (Crossref, OpenAlex and Semantic Scholar; identical text). Three cautions follow. First, the “approximately 2000 µM Ni” sentence in the abstract conflates a bioaccumulation factor (dimensionless) with a solution concentration, and 2,000 µM matches none of the stated treatment levels (1,000 and 3,000 bracket it) — as written it is garbled, and only the full text can say whether the intended figure is a tolerance level or a tissue concentration. We report it as printed, not as read. Second, O. decipiens is described as showing “characteristics of a hyperaccumulator,” which we do not upgrade to a confirmed >1,000 µg g⁻¹ threshold crossing. Third, taxonomic backbones differ: GBIF folds O. decipiens into O. muralis sensu lato, while the POWO treatment accepts it — a naming wrinkle worth knowing before comparing numbers across papers. None of these caveats touches the central, cleanly supported finding: on the soils where it actually lives, the Pirin endemic keeps nickel out.
Source: Pavlova, Nacheva, Karadjova, Tsonev & Bani (2026), Ecological Research 41(5), DOI 10.1111/1440-1703.70121 (published online 2026-08-31; closed access). Supporting references: Reeves et al. 2018; Bettarini et al. 2020. Every load-bearing claim was verified against the publisher-deposited abstract via three independent vendor retrievals (identical text); the full text could not be accessed, and all tissue-level values are deliberately left unquoted for that reason. Disagreements surfaced and resolved: the abstract’s “2000 µM” figure is internally inconsistent and is reported only as printed; the O. decipiens hyperaccumulator claim is kept at the abstract’s own strength. See /methodology/ for how we source and check analysis pieces.