Diogène et al. 2023 — Risk Assessment Strategies for Contaminants in Seafood (RASCS) (EFSA Supporting Publication EN-8419)

This 154-page EFSA Supporting Publication (EN-8419; Question No. EFSA-Q-2023-00708; approved 27 October 2023) is the final report of the RASCS Partnering Grant, an Article 36 capacity-building project executed from June 2021 to June 2023 by a seven-institution EU consortium under coordinator IRTA (Spain). The project’s purpose was knowledge exchange and harmonisation of risk-assessment practice for contaminants in seafood across the participating member states, not original toxicological or occurrence research. The report is organised around six work packages: WP1 hazard identification, WP2 dietary exposure assessment, WP3 risk-assessment strategies in a changing world and balance with nutritional benefits, WP4 risk/benefit perception and communication, WP5 dissemination and outreach, and WP6 management.

For Heavy Metal Index purposes the report contributes regulatory-context and monitoring-inventory material on the four EU-regulated heavy metals in seafood (Pb, Cd, Hg, As) and the analytical methods used to enforce those limits across six member states. It does not present primary contamination measurements, percentile distributions, or exposure estimates for any heavy metal. The heavy-metals content is concentrated in WP1 (regulatory overview, EU MPL table, limits-of-determination table) and WP2 (member-state monitoring-programme inventory for environmental and market surveillance of bivalves, fish, and macroalgae).

Key numbers

• Project structure: six work packages, eleven objectives (O-1 through O-11), seven partner institutions across six EU member states (ES, PT, IT, BE, DE, FR), project duration 24 months (June 2021 – June 2023), four Training-on-Site (ToS) sessions in 2023 (Rome 16–17 February; Lisbon 28 February – 1 March; Barcelona 11–12 April; Paris 11–12 May) and a final workshop at EFSA Parma in June 2023.
• WP1 compound shortlist: from 43 groups of regulated and emerging/non-regulated compounds initially considered, ten groups were selected for further work in subsequent work packages — toxic metals (organic and inorganic arsenic, cadmium, mercury), mycotoxins (ochratoxin A, zearalenone), polychlorinated biphenyls (PCBs), per- and polyfluoroalkyl substances (PFAS), marine biotoxins (TTX, CTXs), and cyanotoxins (MCs) (p. 12).
• EU regulatory framework cited as binding for seafood contaminants (p. 13): Regulation (EU) 2023/915 (consolidated contaminants MPLs, repealing Reg (EC) 1881/2006), Regulation (EC) 853/2004 (specific hygiene rules for food of animal origin including marine biotoxin limits in live bivalve mollusks, echinoderms, tunicates, and marine gastropods), Regulation (EC) 852/2004 (general food hygiene), Regulation (EU) 2017/625 (official controls).
• Heavy-metal MPL scope rule explicitly stated (p. 13): “current legislation does not provide maximum levels for all seafood groups. While mercury is regulated for all mollusks species, maximum levels for cadmium and lead refer to bivalve mollusks only. Maximum levels for arsenic in bivalve species were not considered in the latest Commission Regulation (EU) 2023/465.”
• PFAS regulatory event (p. 13, p. 16): four PFAS (PFOS, PFOA, PFNA, PFHxS) regulated in the EU from 1 January 2023 in food of animal origin including fish meat, bivalve mollusks, and crustaceans, with maximum levels for both individual substances and the sum of the four; introduced via Commission Regulation (EU) 2022/2388 amending Reg (EC) 1881/2006 and subsequently consolidated into Reg (EU) 2023/915.
• EU MPL table for seafood-relevant contaminants (Table 1, p. 16) — toxic-metal rows quoted verbatim:
  • Cadmium (Cd): 0.05–1.00 mg/kg wet weight; official method ICP-MS; Reg (EU) 2023/915.
  • Lead (Pb): 0.3–1.5 mg/kg wet weight (MPL varies by food product); official method ICP-MS; Reg (EU) 2023/915.
  • Mercury (Hg): 0.3–1.0 mg/kg wet weight (MPL varies by food product); official methods ICP-MS or AAS; Reg (EU) 2023/915.
  • Dioxins and dioxin-like PCBs: 6.5–20.0 pg/g wet weight expressed as toxic equivalents (TEQ); official method GC-MS/MS; Reg (EU) 2023/915.
  • PAHs (benzo[a]pyrene): 2.0–6.0 µg/kg; sum 12.0–35.0 µg/kg; official method GC-MS; Reg (EU) 2023/915.
  • PFAS: 2.0–45.0 µg/kg (sum of PFOS, PFOA, PFNA, PFHxS); official method LC-MS/MS; Reg (EU) 2023/915.
• Member-state limits-of-determination (LOD/LOQ) for the three regulated toxic metals in seafood, in µg/kg, by national reference laboratory (Table 2, p. 17). The table reports values for six member states; for the toxic-metal rows:
  • Cd: DE-Lower Saxony 0.6–3.2 / 1.8–9.5 (range depending on AAS vs ICP-MS); DE-Schleswig Holstein 1.3 / 4.0; DE-Bavaria NA; ES-Catalonia NA / 50.0; ES-Canary Islands NA; FR-Anses 5.0 / 17.0; FR-Ifremer NA / 10.0; IT NA; PT 2.0 / 6.0.
  • Pb: DE-Lower Saxony 5.4–7.4 / 16.1–22.2; DE-Schleswig Holstein 2.7 / 8.0; DE-Bavaria NA; ES-Catalonia NA / 300; ES-Canary Islands NA; FR-Anses 1.5 / 5.0; FR-Ifremer NA / 8.0; IT NA; PT 20.0 / 60.0.
  • Hg: DE-Lower Saxony 1.0 / 3.1; DE-Schleswig Holstein 0.7 / 2.0; DE-Bavaria NA; ES-Catalonia NA / 10.0; ES-Canary Islands NA; FR-Anses 2.5 / 8.3; FR-Ifremer NA / 15.0; IT NA; PT 5.0 / 11.0.
• Analytical methods used by the consortium’s member-state monitoring programmes for chemical contaminants in live bivalve mollusks (Table 4, p. 24): official methods — Cd by ICP-MS; Pb by ICP-MS; Hg by ICP-MS or AAS; methylmercury by AAS. Unregulated trace elements monitored: As by ICP-OES; Cr by AAS or ICP-OES; Cu by AAS or ICP-OES; Mn by ICP-OES; Ni by AAS or ICP-OES; Zn by AAS or ICP-OES. In Schleswig-Holstein (DE) the unregulated toxic metals additionally monitored in market bivalves were aluminium, arsenic, total chromium, nickel, and silver (p. 26).
• Macroalgae monitoring (p. 27, France only respondent): macroalgae used as seaweed or as food supplements monitored for toxic metals Pb, Cd, Hg, As, Ni; publicly funded under the directorate for competition policy, consumer affairs and fraud control; between six and nine samples analysed annually 2013–2015.
• French ALTEO case study (p. 25): 24 mussel samples analysed for 11 toxic metals to assess sanitary impact of the Alteo company’s discharge of approximately 20 million tonnes of bauxite residue (“red mud”) into the Cassidaigne Canyon, north-west French Mediterranean, from 1996 to 2015. (RASCS reports the existence of the study; no concentration values are extracted into this report.)
• French market-monitoring sampling effort for 2019 (p. 26): number of fish/crustacean/cephalopod samples analysed varied between none for marine biotoxins and 310 for toxic metals.
• Spanish market-monitoring sampling effort: Canary Islands analysed 1,909 samples for ciguatoxins in 2020; Catalonia analysed five samples for dioxins and PCBs in 2022.
• Consumer survey (WP4): more than 1,800 consumers across Belgium, Poland, and Spain participated in a contaminant-specific seafood risk/benefit perception survey, with hypothetical-bias mitigation via a complementary real-choice exercise with students.
• WP3 combined-exposure example (p. 41): the report discusses how a low-tier Component-Based Approach grouping that places microcystin, cadmium, lead, and mercury into a single assessment group is “simple and conservative” but that assessment refinement should drop cadmium (not neurotoxic) and retain MCs, lead, and mercury for a neurotoxicity-grouped cumulative assessment focused on children. The report does not propose threshold values; it discusses grouping methodology only.

Methods (brief)

Knowledge-exchange consortium project executed via monthly online meetings (year 1, COVID-constrained) supplemented by physical Training-on-Site sessions, thematic webinars, and a final workshop at EFSA Parma. Two structured questionnaires were the principal data-gathering instruments. (1) The monitoring-programme questionnaire (RASCS Appendix A) was an Excel-format instrument with five sheets covering bivalve mollusks – marine biotoxins – environmental monitoring; bivalve mollusks – chemical contaminants – environmental monitoring; bivalve mollusks – official market monitoring; fish and fishery products – official market monitoring; and other seafood (macroalgae) – official market monitoring. Each topic gathered general programme metadata, sampling-strategy details, analytical methods including LOD/LOQ, data-management practices (FAIR principles), and COVID-19 impact. Sent 13 June 2022; responses received from FR, DE, IT, PT, ES-Catalonia, ES-Canaries between October 2022 and March 2023; no responses from Belgium or the Greek institution contacted. (2) The dietary-surveys questionnaire (RASCS Appendix B) was prepared by Anses, reviewed by BfR and EFSA, and circulated September 2022; it gathered identification, sampling strategy, recruitment, information-collection method (FFQ or recall), and publication/use of national in-house dietary surveys on seafood consumption. Risk-assessment methodology reviewed includes the WHO Human Health Risk Assessment Toolkit (2nd ed., 2021); EFSA’s 2019 harmonised guidance on combined exposure to chemicals from multiple sources; the EFSA 2022 roadmap for New Approach Methodologies (NAMs) in New Generation Risk Assessment (NGRA); Integrated Approaches to Testing and Assessment (IATA); the OECD Adverse Outcome Pathway (AOP) concept (OECD, 2017); and Risk-Benefit Assessment methodologies developed by IPMA, presented and discussed at the Lisbon ToS. The report incorporates findings from the IPCC 2022 climate-change-impact assessment and the EFSA CLEFSA project on climate-change emerging risks (EFSA, 2020).

Limitations

• Strategy/programmatic report by design — no primary heavy-metal occurrence measurements, exposure-distribution estimates, or quantitative dose-response work are presented. Statements about contamination prevalence or levels in seafood are not derivable from this document.
• Monitoring-programme inventory (WP2 Task 2.1) is acknowledged in the report itself as non-exhaustive and constrained by questionnaire non-response: no replies were received from Belgium or the contacted Greek institution; only two of five Spanish institutions responded; French responses required six separate contacts. Task 2.2 explicitly concludes that “the information gathered in response to the questionnaire was too limited to allow a refined analysis and giving any methodological recommendations” (p. 28).
• Limits-of-determination table (Table 2) reports laboratory-reported values from questionnaire responses; values are heterogeneous in basis and not normalised. Where two ranges are reported for Cd or Pb in DE-Lower Saxony, the report attributes the range to AAS vs ICP-MS instrument choice; the table does not specify which value corresponds to which instrument for the analytes other than those flagged with the (****) footnote.
• Member-state coverage is limited to six EU states (BE, DE, ES, FR, IT, PT) plus partial Greek attempt. Other major EU seafood-consuming or -producing member states (e.g., NL, DK, IE, EL fully, HR, MT) are not surveyed. Conclusions about EU monitoring practice are not generalisable beyond the surveyed states.
• The report does not perform an exhaustive contaminants review (stated at p. 81): “the conclusions and recommendations herein do not intend to cover all contaminants and are not based on an exhaustive and universal vision of contaminants in seafood … the conclusions and recommendations are oriented to a short-list of compounds that we consider of interest from our own perspective.”
• Approved as a partnering-grant deliverable by RASCS consortium authorship under Article 36 of Regulation (EC) No 178/2002, not as an EFSA Scientific Opinion adopted by the Authority. The disclaimer on p. 3 explicitly notes the document “cannot be considered as an output adopted by the Authority. The European Food Safety Authority reserves its rights, view and position as regards the issues addressed and the conclusions reached.” Citations should reflect this status (Article 36 supporting publication, not EFSA Scientific Opinion).

Implications

• Certification: not threshold-bearing for HMTc work, but documents the binding EU MPL framework relevant to any seafood HMTc category — Pb, Cd, Hg MPL ranges and bases per Reg (EU) 2023/915 (Table 1) and the bivalve-only scope of the Cd and Pb MPLs versus the all-mollusk scope of the Hg MPL (p. 13). The report’s documentation that arsenic MPLs for bivalves were deliberately not introduced in the latest amendment (Reg (EU) 2023/465) is regulatory-context relevant when HMTc considers iAs or tAs limits in seafood categories. The January 2023 PFAS MPL introduction via Reg (EU) 2022/2388 is similarly relevant context for any HMTc consideration of PFAS in seafood.
• Courses: directly usable in the EU regulatory-affairs module for seafood-category brands. Illustrates how seafood-specific MPLs are split across regulations (toxic metals and PFAS in Reg (EU) 2023/915; marine biotoxins in Reg (EC) 853/2004; official controls under Reg (EU) 2017/625) and how monitoring is implemented at member-state level via national reference laboratories. Useful as a worked example of how analytical-method LOD/LOQ variation across member-state labs (Table 2) translates into practical heterogeneity in MPL enforcement sensitivity.
• App: not applicable. No occurrence data for ingredient-level contamination estimates.
• Microbiome: out of scope. The report’s marine-biotoxin and cyanotoxin material (CTXs, TTX, MCs, MC-LR equivalents) is microbiome-adjacent (toxin-producing dinoflagellates and cyanobacteria) and belongs on WikiBiome rather than HMI.
• Cross-source signal: this is the second EFSA horizon-scanning / strategy document in the corpus alongside efsa2024-emerging-chemical-risks-food-feed. Together with the existing EFSA CONTAM Panel scientific opinions on Hg (efsa-mercury-methylmercury-2012), Pb (efsa-lead-contam-2010), Cd (efsa-cadmium-contam-2009), and As (efsa-arsenic-contam-2009), RASCS provides the 2023 EU regulatory and monitoring snapshot against which member-state seafood contamination studies should be read.

Wiki pages this source may touch

• lead — EU MPL framework for seafood and bivalve-only scope of the Pb MPL (regulatory-context only; no occurrence data).
• cadmium — EU MPL framework for seafood and bivalve-only scope of the Cd MPL (regulatory-context only; no occurrence data).
• mercury — EU MPL framework for seafood including all-mollusk scope of the Hg MPL; ICP-MS and AAS as official methods (regulatory-context only; no occurrence data).
• mercury-methyl — methylmercury monitoring noted as AAS-based in member-state programmes (regulatory-context only).
• arsenic — discussion of arsenic MPL gap in Reg (EU) 2023/465 for bivalves; organic and inorganic As selected in the RASCS WP1 ten-compound shortlist (regulatory-context only).
• arsenic-inorganic — same as above; iAs called out as a separate prioritised compound.
• arsenic-total — total arsenic monitored as unregulated metal in DE-Schleswig Holstein (regulatory-context only).
• nickel — Ni monitored in macroalgae market surveillance in France and as unregulated metal in DE-Schleswig Holstein bivalves (regulatory-context only).
• chromium — total Cr monitored as unregulated metal in DE-Schleswig Holstein bivalves (regulatory-context only).
• aluminum — Al monitored as unregulated metal in DE-Schleswig Holstein bivalves (regulatory-context only).
• seafood — overall scope of the report; EU regulatory snapshot.
• shellfish — scope of marine-biotoxin and chemical-contaminant monitoring under Reg (EC) 853/2004.
• fresh-fish — scope of fish-and-fishery-products market monitoring in WP2.
• canned-fish — within scope of processed-food monitoring discussed in WP2.
• seaweed-kelp-foods — macroalgae market monitoring (France) for Pb, Cd, Hg, As, Ni.
• fish-marine-non-predatory — within scope of fish market monitoring.
• fish-marine-predatory — within scope of fish market monitoring; relevant to mercury bioaccumulation discussion in WP3 cumulative-RA example.
• fish — within scope.
• shellfish — within scope.
• bivalve-molluscs — primary focus of WP2 monitoring inventory.
• molluscs — all-mollusk Hg MPL discussion.
• seafood — overall scope.
• seaweed — macroalgae market monitoring.
• eu2023-contaminants-maximum-levels — Reg (EU) 2023/915, the binding contaminants MPL framework for seafood toxic metals and PFAS.
• eu-reg-2022-617-mercury-fish — related Hg-in-fish MPL framework prior to consolidation in Reg (EU) 2023/915.

Verification notes

• 2026-06-03 — Fresh source-page ingest from raw/manual-fetch/Kimi_Agent_Download Corruption Issue/seafood_papers/04_Shellfish/. No prior wiki page (DOI 10.2903/sp.efsa.2023.EN-8419 not previously ingested). PDF pages 1–28 and 36–46 and 81–85 read for content; pages 47–80 (WP4 consumer perception, WP5 dissemination) and 86–154 (references, annexes, ToS workshop reports) skimmed but not extracted as they are out of HMI heavy-metals scope.
• Three-letter metal vocabulary follows CLAUDE.md Part 14: tAs and iAs both listed because the source explicitly distinguishes “organic and inorganic arsenic” in the WP1 prioritised-compound list (p. 12) and Table 4 distinguishes total As (unregulated metal monitored) from speciation. tHg and MeHg both listed because Table 4 (p. 24) lists methylmercury as a separate official analyte measured by AAS in addition to total mercury measured by ICP-MS or AAS.
• Numerical values in Table 1 and Table 2 transcribed exactly as the source reports them, including the () and (***) footnote annotations on the MPL ranges and on the AAS-vs-ICP-MS instrument-dependent LOD/LOQ ranges. No silent unit conversion. Table 2 PFAS row contains no values for any member state in the source (cells empty) and so is not summarised in Key numbers.
• No brand names extracted to Key numbers. The Alteo company is named in the source as the regulatory-event subject of the Cassidaigne Canyon red-mud discharge (Part 12 Exception 1 — regulatory-event subject; the discharge is a documented public-record environmental event), and named institutions (IRTA, CREDA, IPMA, ISS, Ghent University, BfR, ANSES, Anses, Ifremer, IUSA-ULPGC) are scientific-method/institutional names not contamination-value attributions (Part 12 Exception 2). Instrument vendors are not specifically named in the body of the source beyond generic method labels (ICP-MS, AAS, ICP-OES, GC-MS, GC-MS/MS, LC-MS/MS, LC-UV, LC-FLD, HILIC-MS/MS) so no Exception 2 calls are required.
• Part 2 firewall: report explicitly discusses combined-exposure assessment methodology including a worked example grouping MCs, Pb, and Hg into a neurotoxicity assessment group for children; the source proposes no threshold values, and this page does not derive any. The “Implications” section is restricted to noting what the source contributes as regulatory framing for HMTc consideration, not to proposing threshold values.
• License: EFSA Supporting Publications under Article 36 partnering grants are published under EFSA’s standard CC-BY-ND terms; reproduction of consortium-provided images is restricted per the p. 3 notice (no images extracted into this wiki page).
• 2026-06-03 — Fresh-context audit subagent (verdict REVISE) flagged that Table 4 (p. 24) lists Arsenic under ICP-OES only, not “AAS or ICP-OES” as the original draft stated. Verified independently against the source: Table 4 row for Arsenic shows “ICP-OES” in the methods-for-unregulated-contaminants column only; AAS is not listed for As (AAS is listed for Cr, Cu, Ni, Zn). Correction applied to the Table 4 summary in Key numbers.
• 2026-06-03 — Same audit flagged a citation inconsistency between “Reg (EU) 2022/2388” (Key numbers) and “Reg (EC) 2022/2388” (Implications) for the PFAS-introducing regulation. Verified independently against p. 13 of the source: text reads “Commission Regulation (EU) 2022/2388”. The (EU) prefix is correct. Implications-section citation normalised to “Reg (EU) 2022/2388”.
• 2026-06-03 — Same audit flagged the four ToS workshop date pairs (Rome 16–17 Feb 2023; Lisbon 28 Feb – 1 Mar 2023; Barcelona 11–12 Apr 2023; Paris 11–12 May 2023) as not verified within the page range it read (pp. 1–28, 36–45). False positive: the dates appear in Section 3 “Description of exchanges performed” on p. 81 of the source PDF, immediately preceding Section 4. No correction needed.

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.