Rahmi, Pratama, Fauziah and Mahardika (Universitas Serambi Mekkah, Indonesia) report in Medicine and Plant Letters (2026, 3:174–183) a greenhouse pot experiment that asks a fair question: can ordinary ammonium sulfate — the cheapest nitrogen–sulfur fertilizer in the tropics — do the job of synthetic chelators such as EDTA in lead-assisted phytoextraction? The substrate was authentic artisanal gold-mine tailing (Pb 49.3 mg kg⁻¹, pH 8.70), loaded at 5 kg per pot. Vetiver (Chrysopogon zizanoides) was grown for eight weeks at four ZA doses (0, 12.5, 25.0, 37.5 g pot⁻¹) and in two ecotypes: A₁ sourced from contaminated tailing, A₂ from uncontaminated farmland.
The real result: an optimum dose, and a mine-adapted ecotype
Both headline findings reproduce from the paper’s own tables. Shoot Pb concentration in the mine-adapted A₁ rose from 5.88 mg kg⁻¹ (no ZA) to a peak of 8.91 mg kg⁻¹ at 25 g pot⁻¹, then fell to 7.46 mg kg⁻¹ at 37.5 g; the quadratic fit (Ŷ = −0.0035X² + 0.1871X + 5.614, R² = 0.72) puts the vertex at 26.7 g pot⁻¹ — arithmetically self-consistent, and the ecotype × dose interaction is significant. A₁ out-concentrated A₂ at every dose, by 1.6–3.1×. That second finding is the actionable one and matches a growing literature: locally adapted metallophyte populations constitutively run the phytochelatin and vacuolar sequestration machinery that naive conspecifics lack, and for phytoextraction deployment you want germplasm from the contaminated site, not from a nursery.
The trade-off the title buries
The same dose that maximizes concentration halves the plant. Dry biomass fell from 6.19 g (P₀) to 3.12 g at P₂; height at 56 days fell from 37.0 to 16.0 cm. Multiply the paper’s own numbers and the “optimum” collapses: Pb removed per pot at P₀ is 5.88 mg kg⁻¹ × 6.19 g ≈ 0.036 mg; at P₂ it is 8.91 × 3.12 g ≈ 0.028 mg. The untreated control removes more lead per pot than the optimal amendment. The paper calls the growth suppression “the most ecologically significant growth response” of the study in its Discussion — and simultaneously claims the optimum dose works “without producing toxic effects visible at the whole-plant level.” Both statements cannot be true, and throughout, the paper labels a concentration (mg kg⁻¹) as “uptake”: no quantity in mg plant⁻¹ or mg pot⁻¹ is ever reported.
The mass balance that cannot close
This is where a reader should stop. Post-harvest tailing Pb is reported to have fallen from 49.3 to 14.13–27.80 mg kg⁻¹ — that is 108–176 mg of Pb leaving each 5 kg pot. The harvested shoots of the best treatment contain about 0.028 mg. Plants account for roughly 0.016% of the observed decline — a ~6,000-fold shortfall. Roots, vetiver’s documented main Pb reservoir, were never analyzed, no leachate was collected, and there was no plant-free control. Meanwhile ZA drove tailing pH to 3.83–5.26 with daily watering. The parsimonious explanation for the “missing” lead is acid-assisted leaching out of the pot — mobilized Pb heading for drainage, not for harvest. The paper’s conclusion that the decline “confirms active phytoextraction” is not an interpretation of its data; it is contradicted by them.
Chemistry and statistics red flags
The proposed sulfate mechanism is wrong as written: the displayed reaction (Pb²⁺ + SO₄²⁻ + O₂ + 2H₂O → PbSO₄ + 4OH⁻) is charge-unbalanced (+0 on the left, −4 on the right), invokes O₂ with no redox change, and produces hydroxide — a base — in a system whose entire premise is acidification. Its product, anglesite (PbSO₄, Ksp ≈ 1.6×10⁻⁸), is a classic immobilization phase, not the “soluble PbSO₄” the text claims; at most, minor aqueous Pb–sulfate complexes could form. The correlation matrix (Table 6) does not reproduce from the paper’s own tables: the printed Pb correlations are −0.644 and −0.593, but recomputing from the dose-level means gives ≈ −0.98 and ≈ −0.88, and the flagged p < 0.01 for r = 0.961 at n = 4 is statistically impossible (p ≈ 0.04 at best). Shoot Pb was measured on an F732-S “Cold Atomic Absorption Analyzer” — an instrument line designed for mercury, not lead — with no QA/QC reported. And the “optimal” 25 g pot⁻¹ is 0.5% w/w of soil, roughly 10 t ha⁻¹ equivalent — tens of times an agronomic fertilizer rate.
What this means for assisted phytoextraction
Strip the overclaim and two things survive. First, the dose-response optimum and the ecotype contrast are real within this one experiment: sulfur–ammonium chemistry can trade biomass for shoot concentration, and mine-origin germplasm concentrates more Pb — useful design intelligence for vetiver-based work, at pot scale, on one substrate. Second, the paper is a worked example of how assisted-phytoextraction studies can mistake leaching for extraction. Any amendment that drops soil pH below 4 and mobilizes Pb creates a groundwater pathway; if your evidence is a falling soil concentration with no leachate measurement, no root data and no plant-free control, you cannot tell cleanup from displacement. For scale: what genuine Pb phytoextraction looks like is our record for round-leaved pennycress — foliar Pb up to 8,200 µg g⁻¹ — about 900 times what these vetiver shoots achieved. Until a follow-up study closes the mass balance, the honest summary of this one is on our lead hub: concentration up, biomass down, soil lead unaccounted for.
Source: full open-access publisher PDF, read in full; every number above verified verbatim and all arithmetic independently re-derived by three separate reviewers. Selection, sourcing and verification methodology: /methodology/.