Lu et al. 2025 — Profitable chrysanthemum phytoremediation of Cd-contaminated farmland in Zhejiang

Lu and colleagues screened 23 ornamental chrysanthemum cultivars over two growing seasons for their ability to remove cadmium from contaminated farmland in eastern China, then tested a single high-performing cultivar against bamboo-vinegar and EDTA-Si chelator treatments. The source is primary field evidence for phytoremediation and supply-chain mitigation: the trial site is the kind of Cd-contaminated farmland that drives rice exceedances in the region, and the ornamental crop is positioned as an income stream that lets remediation pay for itself without entering the food channel.

Key numbers

Field site and contamination baseline

The 1 ha trial site sits at 28°56’30”-28°56’50” N, 120°07’30”-120°07’40” E in a hardware-processing industrial town in Zhejiang Province with a subtropical monsoon climate. A 2020 survey of the farmland found soil Cd content of 0.32 mg/kg, above the 0.3 mg/kg threshold of the Chinese “Soil Environmental Quality Risk Control Standard for Soil Pollution Risk Control of Agricultural Land (Trial)” (GB 15618-2018). The exceedance rate of rice samples grown on the site was nearly 30%. The authors attribute Cd pollution to sewage leaching from surrounding hardware-processing enterprises.

During the trial, rhizosphere soil Cd ranged 0.47-0.96 mg/kg (mean 0.65 mg/kg, n=23 plots), so 100% of the in-trial soil samples exceeded the 0.3 mg/kg standard.

Single-cropping screen, 23 cultivars, 2021

Six cultivars accumulated more than 0.6 mg Cd per plant per season: in descending order, marigold, ZS-gh, QJ-ht, JL-cj, QJ-hg, and JL-dc. The seven cultivars with enrichment coefficients above 8 were, in descending order, QJ-hg, QJ-dh, JL-yg, JL-cj, ZS-gh, QJ-bh1, marigold, and JS-hj. Nine cultivars completed remediation in five or fewer stages: ZS-gh, JL-dc, JL-cj, marigold, QJ-ht, QJ-hg, ZS-hy, QL-xh, and QJ-dh.

The remediation-time equations (paper §2.4) are:

• M_soil = S_soil × h_soil × ρ_soil / 1000
• M_Cd = (Cd_soil − 0.3) × M_soil
• C_Cd = M_c × Cd_c / 1000
• T = M_Cd / C_Cd

For a 400 cm² per-plant footprint, 20 cm soil depth, and a soil density of 1.1 g/cm³, this gives the 4-13 stage range above.

Multiple-cropping cohort, 14 cultivars, 2022

Of the 23 cultivars, 14 grew to full bloom after overwintering: ZS-zc, L-xf, QX-yz, QJ-hg, JH-hj, QJ-tp, QJ-bh2, QJ-dh, JL-yg, ZS-gh, JL-dc, JL-cj, JL-xy, and QJ-ht. Five of these (QX-yz, QJ-hg, QJ-bh2, JL-xy, JL-yg) maintained 2022 biomass at parity with 2021 and are flagged as the heavy-metal-tolerant subset suitable for repeat cropping.

Flower and stalk Cd were measured separately in the 14-cultivar cohort:

Stalks carried roughly two-and-a-half times the Cd of flowers, making the stalk the dominant Cd sink in the harvestable biomass.

Chelator treatments on QX-yz, 2022

The five-treatment chelator test (Table 2, paper §2.2) used the QX-yz cultivar:

EDTA-Si is a silica-immobilized EDTA composite. The authors conclude bamboo vinegar (T1, T2) is the practical chelator: it raises Cd uptake without significantly damaging biomass, and the QX-yz remediation timeline shortens by about one stage when bamboo vinegar is applied. EDTA-Si raises plant Cd content sharply but inhibits growth enough that net per-plant Cd accumulation falls.

Economic estimate

Across five candidate cultivars (QX-yz, QJ-hg, QJ-bh2, JL-xy, JL-yg), Table 3 reports a per-hectare flower yield of about 3.69×10⁶ to 4.16×10⁶ branches per year and a per-hectare net income of about 57.03×10⁴ to 65.09×10⁴ RMB per year. Construction and maintenance run 0.75×10⁴ RMB/ha/yr, fertilizer and pesticide 3.15×10⁴ RMB/ha/yr, machinery 0.525×10⁴ RMB/ha/yr, land rental 1.8×10⁴ RMB/ha/yr. A farmer on 0.1 ha at two crops per year clears 6.17×10⁴ RMB net, above the cited Yongkang City annual income reference of 5.67×10⁴ RMB. The conclusion identifies the practical implication: phytoremediation of Cd-contaminated farmland can self-finance when the crop has independent market value.

Methods (brief)

Twenty-three cultivars were sourced from the Chrysanthemum Base of Nanjing Agricultural University and seeded between 10 March and 10 April 2021. Each of 23 test plots was 12 m² with 0.3 m × 0.3 m film-wrapped ridges and a one-way drainage ditch. Cultivars were planted in three rows at 20 cm spacing, with one cultivar per plot. The 2021 growing cycle ran 10 April-10 October. After flowering the plants were pruned to 10-12 cm stems with 3-4 leaves and overwintered; the 2022 campaign focused on multiple-cropping effects rather than re-planting, with sequential sampling at the flowering stages of each crop cycle.

Three groups of plants and their rhizosphere soils were randomly sampled per plot during 2021 flowering. In 2022 flower, stalk, and rhizosphere soil of the surviving multiple-crop cultivars were sampled. Soil samples came from the 0-20 cm rhizosphere layer; samples were air-dried, ground, and passed through a 2 mm sieve. Flowers and stalks were dried, weighed, and pulverized.

Cd in soil and plant tissue was determined by ICP-MS (Agilent 7500 series, Agilent Technologies, Tokyo, Japan). Samples were pulverized to <100 mesh, digested with HNO₃-HF (3:1) in a microwave system (Mars 6, CEM), diluted to 50 mL with 2% HNO₃, and filtered through 0.45 µm. Detection limits were 0.03 mg/kg for soil and 0.0001 mg/kg for plant tissue. Quality control used GBW-series soil-certified reference materials, with all analyses conducted in compliance with DD 2005-03 (Technical Requirements for Sample Analysis in Ecological Geochemical Evaluation, Trial; China Geological Survey, 2005) and DZ/T 0279-2016 (Analysis Methods for Regional Geochemical Sample; Ministry of Land and Resources, 2016).

Data reduction was done in Microsoft Excel 2019; statistical analysis in SPSS 25; figures in Origin Pro 2022. Significance is reported at p < 0.05.

The Cd species reported throughout is total Cd; no Cd speciation was performed. The remediation-time and accumulation equations are given in paper §2.4 and reproduced under Key numbers above.

Implications

Certification: This paper should not contribute to edible-crop product or ingredient pools. The measured biomass is ornamental chrysanthemum tissue grown on Cd-contaminated farmland and is, by design, an off-channel sink for soil Cd; it is not food, feed, or a cosmetic ingredient. The source supports supply-chain and territorial mitigation context for Cd-contaminated agricultural land, especially the Chinese rice belt where 30% rice exceedance was observed at this trial site.

Courses: Useful case study for how phytoremediation economics work in practice. The Cd-removal performance only matters if the remediating crop has independent market value (ornamental flowers, in this case) — otherwise the long remediation timelines (4-13 growing seasons for the chrysanthemums tested here) make the approach uneconomic. Also a clean example of how a chelator that raises tissue Cd concentration (EDTA-Si) can still reduce total per-plant Cd accumulation when it suppresses biomass growth.

App: Route as Cd-contaminated-farmland and phytoremediation context for cadmium. Do not route chrysanthemum tissue concentrations to food ingredient pages. Cross-reference for upstream-soil-Cd context relevant to Chinese rice exceedance.

Wiki pages this source may touch

• cadmium
• soil-to-plant-transfer

Verification notes

The PDF is the published version-of-record from MDPI Toxics, CC BY 4.0; DOI 10.3390/toxics13050360. Title, author roster, and publication metadata match the title page. The 23-cultivar count and the 14-multiple-cropping count are stated in §3.1 and §3.2 respectively and were verified against Figures 2 and 3.

The biomass and Cd-content ranges quoted above (67.10-166.08 g, 1.97-5.92 mg/kg, EC 2.98-9.84) are taken from the §3.1 text, not estimated from Figure 2 bar graphs. The per-plant Cd accumulation range 0.26-0.83 mg and 4-13 remediation stages are explicit in §3.1 text. The chelator-treatment percentage-point changes are explicit in §3.4 text.

The single-plant maximum of 0.64 mg in QJ-hg appears only in the Conclusions section (paper §4); it is not in §3.1 narrative and is treated as the per-plant single-plant maximum rather than the per-cultivar mean.

Brand firewall: vendor and instrument names (Agilent 7500, Mars 6 CEM, Origin Pro 2022, SPSS 25) appear only in the Methods description and are retained per the Part 12 methods-vendor exception. The chemical product “EDTA-Si” is described as a silica-immobilized EDTA composite purchased from “Jiangsu Fertilizer Industry Co., LTD”; the manufacturer name is retained as a methods-supplier identifier rather than a brand-attribution-to-contamination claim. “Bamboo vinegar” is described generically as a natural organic acid byproduct of bamboo charcoal production; no commercial brand of bamboo vinegar is named.

Cultivar codes (ZS-zc, QJ-hg, JL-yg, etc.) are scientific-trial identifiers from the Chrysanthemum Base of Nanjing Agricultural University and are retained verbatim. Full Chinese variety names (Zhongshanzichen, Qingjianhuanggai, Jinlingyuegui, etc.) are reproduced as printed in Table 1.

No food-channel ingredient or product is named. The “rice exceedance rate of nearly 30%” is a contextual statement about the trial site’s farmland, not a measurement on the chrysanthemums; it is cited under the contamination baseline and not routed to ingredient pages.

Frontmatter matrices use agricultural-soil and plant-tissue, both of which are in the _no_route block of data/evidence/matrix-to-product-map.json. The routing layer correctly ignores them rather than emitting false product routes. No ingredient or product slugs are claimed.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.