Most nickel hyperaccumulators on the map are Mediterranean or Southeast Asian: the Odontarrhena of the Balkans and Anatolia, the Phyllanthus and Rinorea of Sabah and New Caledonia. Madagascar’s ultramafic outcrops, despite a flora that is otherwise a byword for endemism, are comparatively under-surveyed for the trait. A short field study from the University of Antananarivo, Rasolondraibe, Farasoa & Rabesiaka (2026) in the Romanian Journal of Ecology & Environmental Chemistry, screens the herbaceous cover of one Malagasy nickel deposit and reports two species over the line.
The result, stated plainly
The authors sampled 16 herbaceous species at the Valozoro nickel site, digested aerial parts and roots, and measured nickel in each, then computed a bioconcentration factor (BCF, tissue Ni ÷ soil Ni) and a translocation factor (TF, aerial Ni ÷ root Ni). Two species cleared the Baker and Brooks hyperaccumulator threshold of Ni > 1,000 µg g⁻¹ in the aerial parts, with both BCF and TF above 1:
- Senecio cochlearifolius — aerial 1,252 µg g⁻¹ Ni (roots 617), BCF 4.93, TF 2.03.
- Helichrysum phylicaelofolium — aerial 1,154 µg g⁻¹ Ni (roots 357), BCF 4.54, TF 3.23.
A third, Helichrysum aff. gymnocephalum, is an accumulator but not a hyperaccumulator: aerial Ni 348 µg g⁻¹, BCF 1.37, TF 1.25 — it concentrates and translocates nickel, but falls short of the 1,000 µg g⁻¹ bar. The other thirteen species sit below BCF 1.4 and mostly below TF 1.
One detail is worth pausing on, because it changes how much the table proves. The reported bioconcentration factors add little independent weight: back-computing every row shows the authors applied a single assumed soil nickel — roughly 250 µg g⁻¹ — to all sixteen species (1,252 ÷ 4.93 ≈ 1,154 ÷ 4.54 ≈ 348 ÷ 1.37 ≈ 254), even though no soil was measured and the plants grew at different spots. So each BCF is just that species’ shoot concentration rescaled by the same constant; it carries nothing the shoot value does not, and that ~250 µg g⁻¹ is our inference from the ratios, not a figure the paper states. The load-bearing signals are the shoot concentrations themselves and the translocation factors (shoot nickel above root nickel), which come straight from the tissue data.
Why this is not a footnote
The interesting pattern is the family. Both new records are Asteraceae — and Asteraceae is one of the lineages that keeps re-inventing nickel hyperaccumulation on serpentine. Our database already holds the South African Senecio coronatus, a textbook facultative nickel hyperaccumulator (foliar Ni to ~5,700 µg g⁻¹) whose serpentine populations accumulate while its non-serpentine ones do not, alongside the obligate Berkheya accumulators B. zeyheri (to ~12,860 µg g⁻¹) and B. coddii. A Malagasy Senecio clearing the same bar is a second, geographically independent instance of nickel hyperaccumulation in the genus — the kind of convergence that makes Senecio and the Asteraceae worth systematic screening rather than opportunistic sampling. For agromining, which today leans almost entirely on Odontarrhena chalcidica (foliar Ni up to ~22,000 µg g⁻¹, about 2.2%) as its European workhorse, a wider native germplasm pool matters: a tropical crop already adapted to Malagasy ultramafic soils and climate is worth more there than a transplanted Mediterranean one — provided the trait survives the scrutiny these two single-sample records have not yet had (see below).
The skeptic’s caveat — these are candidate records
Read the two numbers as reported candidate records, not as settled additions to the canon, and the reasons are structural. This is a field survey, and the data as presented read as single samples per species — one plant’s shoot chemistry at one moment, not a replicated population mean. Field foliar nickel is also the measurement most vulnerable to surface soil contamination: on a dusty ultramafic mine site, adhered mineral particles can inflate an apparent “tissue” nickel. The authors did wash their samples in distilled water, which helps — but a distilled-water rinse (no acid wash, with analysis by flame AAS rather than ICP-MS) does not reliably strip Ni-rich serpentine dust, so this is a reduced but not eliminated risk, not an omitted control. And it bites hardest at the margin: 1,252 and 1,154 µg g⁻¹ sit only ~25 % and ~15 % above the 1,000 µg g⁻¹ line — close enough that ordinary sampling, residual dust, or analytical error could move either species back across it in a repeat. The paper also carries small internal inconsistencies — one non-hyperaccumulator species appears under two different names (Helichrysum bracteiferum in the table, H. faradifani in the discussion) and its tabulated values do not fully reconcile between the two — a reminder to treat the taxonomy, and the lower-ranked rows, as provisional. None of this makes the records wrong; it makes them hypotheses to confirm — with washed tissue, replicated plants, and ideally a reciprocal-soil test of the kind that established S. coronatus as facultative — before either species is booked as a hyperaccumulator or a crop.
The conservation tension the survey embodies
There is a quieter implication. The motivation for screening the Valozoro flora is that it sits on an exploitable nickel deposit — and Madagascar’s ultramafic areas are both biodiversity refugia rich in narrow endemics and, increasingly, nickel-mining targets. A survey that turns up metallophytes on land valued for its ore is the whole phytoremediation dilemma in miniature: the same nickel that makes these daisies useful for agromining or mine-site rehabilitation is the reason the habitat that produced them is under pressure. If serpentine-endemic hyperaccumulators are to become agromining stock, their wild populations are also the seed source and the conservation stake — worth documenting and protecting before extraction reshapes the outcrop, not after.
What to take from it
One open-access field survey has nominated two Asteraceae as Malagasy nickel hyperaccumulators, one of them a Senecio echoing a known serpentine accumulator on another continent. That is a genuine lead for both agromining germplasm and serpentine biogeography — and, resting on single field samples cleaned only by a distilled-water rinse, a lead to confirm, not to bank. The most useful next step is unglamorous: return, wash, replicate, and test the soil dependence.
This analysis is grounded in the primary source, which is gold open-access (CC BY): Rasolondraibe, Farasoa & Rabesiaka (2026), Romanian Journal of Ecology & Environmental Chemistry 8(1):151–158, doi:10.21698/rjeec.2026.111. We retrieved the full-text PDF and extracted every figure above verbatim from its abstract and Table 1, then cross-checked each number — the two shoot concentrations, the BCF and TF values, and the internal arithmetic — against the source across three independent model vendors (Claude, GPT and Kimi) plus an independent web re-retrieval. Species figures for Senecio coronatus, Berkheya zeyheri and Odontarrhena chalcidica follow the compilation in our database; see our methodology for how records are verified.