Camurati 2019 — Arsenic in edible macroalgae (critical review)

Camurati and Salomone reviewed the literature to 2019 on arsenic in edible marine macroalgae, integrating arsenic speciation and toxicity, the behaviour of arsenic species during digestion and cooking, analytical/CRM challenges, and current legislation, with a view to informing future regulation as seaweed consumption grows (including in South America). This is a secondary source: it reports no primary measurements, and its quantitative statements are syntheses of cited primary studies. Its principal value for the index is the synthesized picture it confirms — that inorganic arsenic in seaweed is generally low but with species-specific high exceptions (hijiki and Laminaria digitata), that brown algae carry both higher total arsenic and a higher inorganic fraction than red or green algae, and that cooking and digestion can change arsenic bioaccessibility and speciation in ways current total-As/iAs limits do not capture.

Key numbers

This is a review; the figures below are the review’s syntheses of cited primary literature, not new measurements, and should be treated as leads/context rather than poolable occurrence data.

• Total arsenic in macroalgae spans roughly 0.1–250 mg/kg dw across species (review of Ma et al. 2018).
• Inorganic arsenic “rarely exceeds 1 mg/kg dw in seaweed” (citing Almela 2002, Díaz 2012, Morrison 2014, Pétursdóttir & Gunnlaugsdóttir 2019), with notable exceptions hijiki (Hizikia fusiforme) and Laminaria digitata (citing Ichikawa 2006, Mise 2019, Ronan 2017).
• Brown algae (Phaeophyta) contain higher total arsenic, and a higher proportion of inorganic arsenic relative to total, than red (Rhodophyta) or green (Chlorophyta) algae (citing Desideri 2017, Ma et al. 2018).
• Organic arsenic can reach ~89% of total in seaweed (citing Cubadda et al. 2017); all organic As species except arsenobetaine are considered potentially toxic.
• Seawater arsenic typically 0.5–2 µg As/L; total As in macroalgae 0.1–250 mg/kg dw; WHO estimate of total dietary As intake 20–300 µg/day.
• Digestion/cooking transformation (cited primary findings): ~85% of inorganic arsenic in hijiki is released into gastrointestinal juices and thus bioaccessible (Brandon 2014); in vitro bioaccessibility of total As from macroalgae ranged ~12–80% across studies; cooking can raise total-As availability (e.g., Porphyra 87→106%; H. fusiforme As(V) rising 8.65→41.7 µg/g, Almela 2005); 34–71% of As can be removed into cooking water (García-Sartal 2012).
• Exposure context: Japanese children and pregnant women may be exposed to high inorganic arsenic via daily hijiki consumption (Mise 2019); a 3% hijiki diet caused arsenic accumulation and poisoning signs in rats (Yokoi & Konomi 2012).

Legislation summary (as compiled by the review):

• Australia & New Zealand: maximum inorganic As in seaweed 1 mg/kg dw (ANZ Food Standards Code).
• France and USA: maximum inorganic As 3 mg/kg dw in algae-based products.
• Argentina (Argentine Food Code): no specific maximum for As in macroalgae; only a general 1 mg/kg total-As for solid foods, and 0.3 mg/kg inorganic As for leafy vegetables / 1 mg/kg for fish/shellfish/bivalves — which the authors argue is inappropriate for seaweed.
• General (the review does not name the EU or Commission Regulations): existing food and drinking-water limits refer only to total or inorganic arsenic; regulation of maximum As levels in edible seaweed is limited and principally refers to inorganic forms; the review argues species-specific regulation based on individual As compounds is needed for edible seaweed.

Analytical/CRM note: certified reference materials for arsenic species in seaweed are scarce — SRM 3232 Kelp Powder (Thallus laminariae, Yu et al. 2018, first to certify arsenosugar species), Ulva lactuca BCR-279 (total As), NMIJ CRM 7405-a hijiki (t-As + As(V)), and NIES CRM 09 Sargassum fulvellum (t-As) are the commonly used materials.

Methods (brief)

Narrative critical review. The authors searched SCOPUS using keywords (marine algae, seaweed, macroalgae combined with arsenic, inorganic arsenic, and other arsenical compounds), covering literature published up to 2019. They organised the synthesis around arsenic speciation and toxicity, arsenic in macroalgae, transformation during digestion and cooking, extraction and analytical techniques (HPLC-ICP-MS, HPLC-HG-AAS/AFS) and CRM availability, and current legislation. No primary sampling or measurement was performed; all concentrations and percentages are attributed to the cited primary studies.

Implications

• Certification (HMTc): a B-tier secondary source that corroborates, from an independent review vantage, the structural premises of the Category 6 seaweed-kelp-foods row — low background inorganic arsenic across most edible seaweeds with species-specific high exceptions (hijiki and Laminaria digitata), and a brown > red > green ordering of both total arsenic and inorganic fraction. It also assembles the comparative legislative landscape (ANZ 1 mg/kg; France/USA 3 mg/kg iAs in algae products; Argentina no seaweed-specific iAs limit; and the general observation that existing food/water limits refer only to total or inorganic As) useful as regulatory context, and flags that cooking/digestion transformations are not captured by static total-As/iAs limits. Because it is a review, its numbers are leads, not pooling inputs; the underlying primary studies (several already in this folder) carry the occurrence weight.
• Courses: a strong orientation reading on why arsenic speciation, CRM scarcity, and cooking/bioaccessibility matter for seaweed food safety.

Verification notes

• raw_handle MFK_camurati2019 from the PDF filename; raw_path under “raw/Manual Fetch Kimi /June 8 Inorganic Arsenic Seaweed/“. DOI 10.1080/10937404.2019.1672364 confirmed on the article header.
• Evidence tier B and source_type review: this is a critical review (J. Toxicol. Environ. Health Part B is a reviews journal). Per Part 13, review claims are B-tier attributed and are recorded here as context/leads, explicitly not as primary occurrence data for the pooling/extraction pass. sample_n is null (no samples).
• Speciation: the review discusses total As and inorganic As (As(III)+As(V)) plus organic species (arsenosugars, AB, AC, DMA, MMA, arsenolipids); only tAs and iAs are lifted to frontmatter as the index-relevant analytes. No primary speciated values are claimed by this source.
• Units/basis: figures are reproduced as the review states them (mg/kg dw, µg/g, µg/L, %), each attributed to its primary source; no conversions applied. These are not to be pooled.
• jurisdictions [AR]: although the review synthesizes global literature, it is Argentine-authored, devotes substantial analysis to the Argentine Food Code, and is explicitly motivated by the growing South America / Argentina seaweed market, so AR is its regulatory/market lens. (Set to AR rather than [] both to reflect that framing and to clear the non-blocking routing advisory that empty jurisdictions triggers.) The multi-jurisdiction legislation summary (ANZ, France, USA, EU) is content, captured in the body, not additional jurisdiction tags.
• Brand firewall: not engaged (review; no products or brands named).
• matrices [edible-seaweed, macroalgae, kelp, dry-weight] per corpus convention; kelp retained because the review treats Laminaria prominently. Kept broad.
• Instrument/CRM names (SRM 3232, BCR-279, NMIJ 7405-a, NIES 09) retained in the analytical note as permitted scientific reporting.
• Audit subagent (2026-06-08, fresh-context) returned REVISE; all numbers, speciation labels (tAs/iAs; As(V) not mislabeled as iAs), slugs, and both firewalls verified clean, and the B-tier/review/not-poolable framing and jurisdictions: [AR] confirmed defensible. One ⚠️ applied: the legislation summary had attributed an “EU: Commission Regulations” framing the review never makes (the review’s actual statement is generic — existing food/water limits refer only to total or inorganic As); reworded the legislation line and its Implications echo to the review’s actual unattributed wording. No numeric/speciation corrections needed.
• Cross-source note: this review cites several primary studies also held in this folder (e.g., Almela 2002, Rose 2007, Ronan 2017, Taylor & Jackson 2016, Llorente-Mirandes 2010/2011); cross-source synthesis against them belongs to the Part 9 synthesis pass, not this source page.

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.