Literature

Chamomile crosses the 100 mg kg⁻¹ cadmium line at one dose in a greenhouse pot — and its best crop still removed only 0.18% of the soil cadmium

Kikis, Grammenou, Thalassinos, Petropoulos and Antoniadis (University of Thessaly, Greece) report in Plants (2026, 15(20):3088) a 120-pot greenhouse experiment that separates two things the phytoremediation literature routinely conflates: accumulating a metal in a shoot and removing it from a soil. Chamomile (Matricaria recutita) and safflower (Carthamus tinctorius) were grown for four months in an alkaline loam (pH 7.9, 20% CaCO₃, background Cd 0.05 and Pb 3.57 mg kg⁻¹) spiked with CdCl₂ at 0/50/100 mg Cd kg⁻¹ or PbCl₂ at 0/1500/3000 mg Pb kg⁻¹ — doses chosen to match documented mining hotspots, with a concurrent chloride-only control to rule out the counter-ion.

The result that looks like hyperaccumulation — at exactly one dose

At Cd50, chamomile shoots reached 138.6 mg Cd kg⁻¹ DW, with a shoot bioconcentration factor of 3.04 and a translocation factor of 1.56 — numerically clearing all three legs of the commonly cited Cd screening criteria (shoot >100 mg kg⁻¹, BCF > 1, TF > 1). This in a calcareous soil, where carbonate sorption is supposed to starve plants of Cd. The paper, notably, never uses the word “hyperaccumulator”: the chamomile literature is genuinely split, and one of the papers it cites carries the title “Matricaria chamomilla Is Not a Hyperaccumulator, but Tolerant to Cadmium Stress.”

Why the label would not survive the next dose

At Cd100, chamomile shoots went higher — 208 mg kg⁻¹, BCF 2.18 — but the physiology behind the numbers collapsed: translocation fell to 0.52 (Cd was retained in roots instead, 400.6 mg kg⁻¹), shoot dry biomass dropped from 2.9 g to 0.2 g, and total aboveground Cd uptake fell roughly fivefold, from ≈250 to ≈50 µg pot⁻¹. A concentration threshold cannot see this, because it has no biomass and no dose axis in it: the same plant both “is” and “is not” a hyperaccumulator depending on which side of the dose curve it is sampled on. For calibration, what field-proven Cd hyperaccumulation looks like is Sedum alfredii or Noccaea caerulescens — species that hold shoot Cd in the thousands of mg kg⁻¹ while still producing harvestable biomass.

Safflower’s lead hormesis — real, unexplained, and mostly below ground

Safflower’s headline response ran the other way: at Pb1500 it grew better, with shoot dry weight up 17% (to 4.1 g) and fresh weight up 44% (to 33 g) against the control, reversing to severe suppression at Pb3000 (shoot DW 1.5 g). The +44% is mostly water — dry matter is the +17% — and the mechanism is explicitly not demonstrated: the authors cite melatonin-pathway and CtWD40-6 hypotheses from other studies and state plainly that no physiological measurements were made to support them. Their own chloride-only control (CaCl₂ at the Pb3000 chloride loading) showed no growth effect, so the stimulation is not a salt artifact. What safflower did not do is move Pb upward: TF stayed below 1 at both doses (0.65 → 0.25), with roots out-concentrating shoots — a belowground-retention phenotype suited to phytostabilization, not extraction.

The number that disciplines the rest: 0.18%

Divide the best shoot uptake by the measured, post-harvest soil pool and the framing changes completely. Chamomile at Cd50 pulled ≈250 µg Cd from a pot whose soil held 136.8 mg Cd after equilibration (measured 45.6 mg kg⁻¹ × 3 kg — note this is ~91% of the 150 mg nominally spiked): that is ≈0.18% of the soil Cd pool in one four-month cycle, with total plant biomass of just 1.81 g dry. Safflower at Pb1500 removed ≈300 µg Pb against 3,699 mg in the soil — ≈0.008%. The uptake figures are read from bar charts, not tables, and sit ~10% above the concentration × biomass product, so treat them as approximate; but no plausible rounding moves either figure by an order of magnitude. The paper says this itself, repeatedly and honestly: a high BCF “is not synonymous with substantial bulk soil decontamination,” and meaningful Cd drawdown would take many sequential harvests.

The alkaline-soil surprise worth keeping

The geochemistry did push back, but not enough, and not evenly. DTPA-extractable Cd was 30–36% of pseudo-total at Cd50 and over 53% at Cd100; for Pb the extractable share rose from 44–47% at Pb1500 to 51–53% at Pb3000 — the pattern of high-affinity retention sites progressively saturating as loading increases. So the alkaline matrix slowed metal partitioning but did not prevent a large chelator-accessible pool from forming at hotspot doses. That has a monitoring corollary: in calcareous contaminated land, the fraction of metal that is available can grow with the total, which is exactly the regime where phytostabilization choices (safflower, root retention) matter more than extraction ones.

Bound the claim and it is useful

This is one soil, one season, one harvest, metal salts spiked into initially clean soil and incubated for only 30 days — a screening design, and the authors label it as such. What survives the bounding: (1) chamomile expresses strong, dose-dependent Cd accumulation and translocation even under alkaline geochemistry, worth repeated-harvest evaluation as the authors suggest; (2) safflower’s Pb response is a tolerance-plus-root-retention phenotype, worth long-term phytostabilization assessment — and its hormetic growth peak is a reminder that “thrives on contaminated land” and “extracts the metal” are different claims; and (3) the 0.18% figure is the sanity check every BCF-led screening study should be forced to report. One caution the paper underlines implicitly: chamomile is a consumed medicinal plant, so biomass rich enough to score 138.6 mg kg⁻¹ Cd is a disposal liability, not a product. That trade-off — productive use versus extraction versus food-chain risk — is the real management question on our cadmium hub and lead hub.


Source: full open-access text (MDPI, CC BY), read in full; every number verified verbatim and all arithmetic independently re-derived by three separate reviewers; retrieval and cross-check notes per /methodology/.

Primary source: https://doi.org/10.3390/plants15203088

← All news & analysis