The assisted-phytoextraction literature is full of papers that report a successful intervention as a bigger bioconcentration factor. Quronfulah, El-Aidarous and colleagues at Umm Al-Qura University (Saudi Arabia) have published the useful counter-example: a greenhouse trial on genuinely nasty soil in which the strongest chelators were precisely the treatments whose total metal uptake fell, because they gutted the crop (J. Saudi Soc. Agric. Sci. 25(7), 2026).
The soil is the real thing
The pots were filled with topsoil from the Mahd Ad’Dahb gold mine: sandy loam, pH 7.66, 3.76% organic matter, and total metal loads of 25.5 mg kg⁻¹ Cd, 597 Cu, 1,960 Fe, 420 Pb and 797 Zn, of which the DTPA-extractable fractions were 11.2, 27.8, 45.9, 32.5 and 79.3 mg kg⁻¹ respectively. That is a heavily contaminated substrate, not a spiked sand. Eight treatments crossed three soil-applied chelators — EDTA, DTPA and citric acid, each at 5 mmol kg⁻¹ applied 70 days after sowing — with or without a foliar spray of the auxin IAA (10 mL per pot at 35 days). Castor (Ricinus communis), a fast-growing oil and bioenergy crop rather than a hyperaccumulator, was the test plant.
The trade-off, measured
All three chelators significantly raised BCF values for all five metals. But they did not do it equally, or freely. The synthetic aminopolycarboxylates EDTA and DTPA significantly reduced root length and plant height — the paper reports declines of up to 15.6%, 25.4%, 15.9% and 14.9% across the two chelators — and with that biomass loss came the result that matters: “applying DTPA and EDTA increased HMs absorption but decreased biomass; therefore, total HMs uptake was reduced in the castor oil crop.” Tissue concentrations went up while the kilograms of metal per pot went down. The paper’s own correlation analysis names the mechanism: plant dry weight is significantly negatively correlated with tissue metal content (root dry weight vs. root metal content, r = −0.576 to −0.338), while total uptake of every metal increases with dry biomass. Concentration × biomass is the product being sold, and the synthetic chelators were sacrificing exactly the second factor.
Citric acid behaved differently. Alone, it raised shoot metal contents by 2.7-, 1.4- and 1.1-fold for Cd, Cu and Fe/Pb/Zn respectively, without the growth penalty. And paired with the IAA spray it was the best treatment in the experiment: translocation factors of 0.66, 0.25, 0.46, 0.36 and 0.49 for Cd, Cu, Fe, Pb and Zn, versus 0.21, 0.24, 0.40, 0.29 and 0.40 in controls, and total uptake reaching 17.8, 253.7, 362.9, 202.4 and 407.5 mg plant⁻¹ for the five metals, against 6.4, 177.9, 276.9, 168.6 and 353.6 mg plant⁻¹ in unamended controls — roughly a tripling of per-plant cadmium removal (6.4 → 17.8 mg).
Why this is the non-obvious result
The standard assisted-phytoextraction story — mobilize the soil pool with a strong chelator, watch shoot metal climb — has always had an implicit assumption that the plant would keep growing. This trial puts numbers on how badly that assumption can fail at agronomically realistic chelator doses, and shows the cheap alternative: a readily biodegradable organic acid that mobilizes metals mildly, plus a growth regulator that pushes biomass, so that concentration and yield rise together. For the growing niche of “phytoremediation crops you can sell” — castor’s seed oil is the commercially attractive part, with the metal harvest a remediation by-product — a treatment that preserves yield is the only kind that is economically legible at all. It is also the conservative choice environmentally: aggressive synthetic chelators applied to a field soil risk mobilizing metals that no root captures, moving Cd and Pb toward drainage rather than toward the harvest.
The honest bounds
This is one greenhouse pot trial, on one soil, with metal data from n = 3 replicates, and castor is a non-hyperaccumulator: even the best treatment moved a modest fraction of a severely contaminated soil’s metal pool into one season’s biomass, so the remediation horizon implied here is long and multi-crop. The authors’ own framing — a candidate approach “suitable for use with bioenergy crops” on mining soils — is appropriately narrower than a clean-up claim. Note also that the paper defines total uptake in µg pot⁻¹ in its methods but reports the figure values as mg plant⁻¹; the numbers are quoted above as printed. And a trial like this cannot see the leaching risk it warns against — that needs lysimeter or field work, ideally on exactly this soil. For context on the metal this crop moved most effectively, see the site’s cadmium and lead hubs; the contrast class — plants whose tolerance is constitutive enough to carry extreme tissue metal without the biomass penalty — is what separates a bioenergy crop with a remediation bonus from a true hyperaccumulator-driven clean-up.
Source: Quronfulah, El-Aidarous, ALgthami, Nahari, Al-Malki, Melebari & Osman (2026), Journal of the Saudi Society of Agricultural Sciences 25(7), DOI 10.1007/s44447-026-00196-y (gold open access, CC-BY; retrieved 2026-09-22 via the article’s full text and Table 1). Every load-bearing number — the soil totals and DTPA extracts, the 15.6/25.4/15.9/14.9% growth penalties, the TF and BCF increases, the fold-changes in tissue metal, the correlation coefficients, and the 17.8/253.7/362.9/202.4/407.5 vs 6.4/177.9/276.9/168.6/353.6 mg plant⁻¹ uptake pair — was verified verbatim by three independent reviewers on three different model vendors, each retrieving the primary source independently; no numeric discrepancies were found, and the reviewers’ flags (the µg pot⁻¹ vs mg plant⁻¹ unit inconsistency, the unmapped order of the four growth-penalty percentages, and the small n) are disclosed in the text above. See /methodology/ for how we source and check analysis pieces.