Published online 2026-09-26 in Ecotoxicology and Environmental Safety (vol 324, article 120835), the study by Chen, Lai, Cui and colleagues (Guilin University of Technology and Zhejiang University) asks a question that matters more with every remediated landscape: when a contaminated site is cleaned up, does the plant population that evolved there keep what the contamination taught it? The experimental design is clean. Two ecotypes of Leersia hexandra Swartz — a grass long reported as a chromium hyperaccumulator from Chinese electroplating-region field studies — were collected from a historically contaminated site that has since been remediated, and from uncontaminated cropland, then grown in hydroponics under combined Cu(II) and Ni(II) stress. The site history made the comparison possible in the first place: the formerly contaminated population has lived without its selective pressure for years, so any trait difference that survives is a measure of evolutionary lag, not active selection.
The numbers, and the organ that does the defending
The contaminated-site ecotype accumulated significantly more metal than the cropland ecotype (p < 0.05 for overall accumulation capacity), with aboveground tissue differences of up to 17.6-fold for Cu(II) in stems and 23.5-fold in leaves; Ni(II) tracked the same pattern, though the abstract reports that trend only qualitatively. The physiological signature is the interesting part. The stronger accumulator did not defend its leaves harder — it defended its roots harder, with peroxidase (POD) activity 5.6-fold higher and catalase (CAT) 14.0-fold higher than the cropland ecotype, while leaf antioxidant enzyme activities stayed comparatively low. Read as a mechanism, that is a compartmented strategy: keep the redox front line at the root surface, where the metal enters, and invest instead in moving metal onward into stems and leaves — exactly the shoot-allocation phenotype a phytoextraction crop needs. The abstract does not give absolute tissue concentrations, so it is not possible to check the result against the >1000 µg g⁻¹ dry-mass hyperaccumulation threshold for Cu or Ni; what is demonstrated is a between-ecotype contrast in allocation and root defense, not a confirmed hyperaccumulation record for either metal.
The non-obvious implication: remediated sites are germplasm archives
The authors’ own framing is that “historically selected metal-adaptive traits in L. hexandra may persist to some extent after environmental pressure declines.” The practical corollary cuts against current practice. Germplasm prospecting for phytoremediation and agromining concentrates on sites that are still contaminated, on the reasonable assumption that selection is strongest there. This result says the archive extends further back in time than that: a remediated electroplating site or closed mine can retain high-value populations whose trait advantage survives years of relaxed selection — and such sites are precisely the ones being redeveloped, graded over or restored with non-local plant material, so the archive has an expiry date measured in land-use decisions rather than generations. There is also a conservation mirror-image: if traits persist this long, then “already remediated” is not a reason to deprioritise a metallophyte population for protection or collection.
The genuine risks, stated plainly
Three cautions belong in the same breath as the opportunity. First, the accumulation capacity of the contaminated-site ecotype was lower than the historical maximum reported for the same habitat — the authors read this as partial trait erosion under relaxed selection, which implies the persistence is real but leaky, and no buyer of this germplasm should assume the historical peak performance. The abstract does not quantify the decline or cite the historical source, and a plausible candidate for that source is the co-authors’ own earlier work on this grass’s chromium accumulation, so the “historical maximum” benchmark should be treated as internal to the group until the full text says otherwise. Second, this is one population contrast, in hydroponics, over what is necessarily a short renewed-exposure assay; fold differences of this size in a controlled solution culture routinely compress in soil. Third, the firmly established hyperaccumulator credential for L. hexandra is Cr — the 2007 Chemosphere report of up to 2,978 mg kg⁻¹ leaf Cr near an electroplating factory — while its Cu/Ni hyperaccumulation rests on softer, partly non-primary literature; this paper strengthens the tolerance and allocation case for Cu and Ni without settling threshold status.
Where it connects here
The species is not yet in this site’s database; the honest links are the element hubs the work touches — /metals/copper/ and /metals/nickel/ — and the broader pattern of ecotypic differentiation among metallophytes that our recent serpentine-genome and ecotype pieces track. If the Cu/Ni trend holds in soil trials and the shoot numbers clear threshold on an absolute basis, L. hexandra would graduate from a single-metal (Cr) curiosity to a genuine multi-metal candidate for Cu–Ni co-contamination — the signature waste stream of electroplating and mining districts in southern China, where this group works.
Source: Chen M, Lai C, Cui W, Yang X, Zhang X, Jiang P, Liu J, Qiu H (2026), Ecotoxicology and Environmental Safety 324:120835, DOI 10.1016/j.ecoenv.2026.120835 (e-pub 2026-09-26). The article is paywalled and the ScienceDirect full text is not machine-readable from our network, so every claim above is bounded to the peer-reviewed abstract as indexed by Europe PMC (PMID 42800339): the two ecotypes, the 17.6- and 23.5-fold Cu stem/leaf contrasts, the POD 5.6-fold and CAT 14.0-fold root-enzyme contrasts, the qualitative Ni trend, and the below-historical-maximum finding. Each of those numbers was verified verbatim by three independent reviewers on three different model vendors, each retrieving the primary record independently; no discrepancies were found, and the per-organ attribution of the p < 0.05 significance statement is left conservative deliberately. See /methodology/ for how we source and check analysis pieces.