Nielsen & Nathan 1975 — Heavy metals in 13 New Zealand edible mollusc species (national survey)

Nielsen and Nathan (Fisheries Research Division, Ministry of Agriculture and Fisheries, Wellington) analysed cadmium, lead, copper, mercury, zinc, and iron in 203 numbered samples drawn from 199 sites around the New Zealand coast, covering 13 edible mollusc species (12 bivalves plus the gastropod Haliotis iris), with at least 14 average-sized specimens combined and homogenised per sampling point so each reported value represents the local population mean. Analysis used atomic absorption spectrophotometry (AAS) on dry-ashed tissue dissolved in 3 M HNO₃, with mercury measured separately by the Stux & Rothery 1971 cold-vapour method (volatility precludes the dry-ashing approach). All concentrations are reported in µg per g wet weight (= mg/kg WW = ppm WW) on whole soft tissue, with three exceptions: scallop Pecten novaezelandiae values are stomach-only by default because the stomach uniquely accumulates cadmium (Brooks & Rumsby 1965), and gonad-only or adductor-only sub-samples were taken for a small number of P. novaezelandiae and M. edulis samples (footnoted in the Appendix). Wide site-to-site variation dominates the dataset: oyster cadmium near Brampton Shoals (Bay of Islands) was nearly twice that at Wairoa Bay less than 1 km distant; rock oyster Crassostrea glomerata at one Bay of Plenty sewer-outfall sample (Sample 61, Omokoroa) reached 21 µg/g Pb (excluded from the species average as abnormal but retained in the range); copper in C. glomerata at the defunct Kawau Island copper mine averaged 380 µg/g and remained at 305 µg/g 20 m from the mine and 165 µg/g 40 m away. Scallop stomach cadmium reached 329 µg/g in one Tasman Bay sample (Sample 164), with stomach-tissue averages of 137 µg/g across all P. novaezelandiae samples — orders of magnitude above adductor (0.51 µg/g) and gonad (1.5 µg/g) tissue from the same animals. The authors interpret the broad pattern as reflecting New Zealand’s geologically young volcanic landmass (mineral-rich runoff into the sea), supplemented by point-source signals from sewer outfalls, urban lead deposition (vehicle exhaust and lead-based roofing paint, especially around Wellington), and superphosphate-fertilizer cadmium reaching Tasman Bay via aerial crop-dusting runoff (Williams & David 1973 cited).

Key numbers

Units: µg per g wet weight (= mg/kg wet weight = ppm wet weight) throughout, on whole soft tissue unless otherwise noted. The Appendix footnotes: * = gonad only; † = stomach only; ‡ = adductor muscle only; § = foot only. The dry-ashing procedure (12 h at 100°C, then 12 h at 450°C) is sample preparation, not a basis change — values remain wet-weight. Mercury was analysed by cold-vapour after HNO₃/H₂SO₄ reflux because the dry-ashing temperature volatilises mercury. AAS recovery: Cd 100±2%, Zn 99±2%, Pb 93±4%, Fe 84±3%, Cu 100±1%, Hg 100±3% (results in Appendix not corrected for recovery). Lead values below 0.5 µg/g are flagged by the authors as questionable (no hydrogen-lamp scattering correction), but were included to indicate probable trace presence.

Cadmium averages by species (whole soft tissue, µg/g wet weight)

• Ostrea lutaria (3.9 µg/g) and P. novaezelandiae stomach (137 µg/g) accumulate cadmium to an unusually high degree. Anomia walteri (2.00) and C. glomerata (1.3) also run elevated. Fig. 4 histograms: values of >2 µg/g Cd are uncommon in C. glomerata; in O. lutaria Cd is widely distributed at 1-6 µg/g; in P. canaliculus only a few samples had >0.5 µg/g Cd.
• The scallop stomach-vs-edible-tissue contrast (137 µg/g stomach vs 0.51 µg/g adductor muscle and 1.5 µg/g gonad) is the canonical example: the stomach is the bioaccumulating organ but is not normally eaten; the adductor and gonad (which are eaten) do not concentrate cadmium to dietary-risk levels.
• Tasman Bay scallop stomach Cd map (Fig. 3): the largest circles represent 329 µg/g; the highest cadmium falls in the southwestern Tasman Bay area, with Nelson City sewage outfall flowing into the bay but no specific effluent assay performed.

Lead averages by species (whole soft tissue, µg/g wet weight)

• Sample 61 (C. glomerata near a Wellington-area sewer outfall, “Omokoroa sewer outfall” in the Appendix listing): Pb 21 µg/g; explicitly excluded from the C. glomerata average as an abnormal sample but retained in the species range. The text further notes “samples of some species from a few locations had quite high lead values.”
• Perna canaliculus (green-lipped mussel) lead is roughly twice that in other species (except Chione stutchburyi cockle, which is less commonly eaten). The species “seems to have a predilection for accumulating lead” and runs higher in the vicinity of large cities. The Wellington-area sub-study (Fig. 2): the largest plotted circle is 7.8 µg/g; populated areas shaded; values clearly higher near the city than at offshore stations. The authors attribute the Wellington Harbour pattern to lead-based roofing-paint leaching via storm drains and to automotive lead exhaust deposition.
• Fig. 5 histograms: most C. glomerata samples were under 1 µg/g Pb; most P. canaliculus samples were over 1 µg/g Pb; the P. canaliculus lead distribution is wider, “which may indicate a greater physiological ability to accumulate or tolerate lead.”

Mercury averages by species (whole soft tissue, µg/g wet weight; cold-vapour method)

• All six species averaged “much less than 0.50 µg/g” total mercury. Fig. 6 (mercury histogram for C. glomerata): regular distribution with greatest frequency 0.1-0.2 µg/g.
• Mercury was analysed only on a representative subset because of the volatility-driven need for the cold-vapour method rather than the dry-ashing AAS used for the other metals. The Appendix shows blanks (”---”) for Hg in most samples, indicating “not analysed” rather than “below detection.” Speciation flag: total mercury (tHg). Methylmercury was not measured.

Zinc averages by species (whole soft tissue, µg/g wet weight)

• C. glomerata zinc (337 µg/g average) is five times the O. lutaria average and more than ten times all other species. Fig. 7 histograms show most C. glomerata samples >100 µg/g and most O. lutaria samples <100 µg/g. The high oyster zinc-accumulating capacity matches published international experience (Thrower & Eustace 1973 cited Tasmanian oysters at >9000 µg/g in polluted waters).
• C. glomerata averaged about three times the cadmium of O. lutaria but about ten times the zinc; the authors note this as either a physiological-concentration difference or a difference in environmental availability between the two species’ typical ranges.

Copper averages by species (whole soft tissue, µg/g wet weight)

• Kawau Island defunct copper mine (Samples 38-40) point-source signal: C. glomerata at the mine averaged 380 µg/g Cu; 20 m from the mine 305 µg/g; 40 m from the mine 165 µg/g. The text notes “slightly greenish tissue, and green deposits inside the shell” in some Kawau oysters — historically attributed to copper exposure (cf. Galtsoff 1966 on green-coloured American oysters).

Iron

Iron was found up to 280 µg/g in Perna canaliculus. The authors decline to present iron species-averages “since iron is not a toxic heavy metal” and refer the reader to the Appendix for site-level iron data.

Selected high-value site samples worth noting (from Appendix 1)

• Sample 61 — C. glomerata, “Omokoroa sewer outfall” (Bay of Plenty): Pb 21 µg/g (the abnormally-high sample excluded from species average).
• Sample 38 — C. glomerata, “Copper Mine” (Kawau Island): Cu 380 µg/g.
• Sample 39 — C. glomerata, “20 m E of copper mine”: Cu 305 µg/g.
• Sample 40 — C. glomerata, “40 m E of copper mine”: Cu 165 µg/g.
• Sample 47 — C. glomerata, “Rocky Bay” (Waiheke Island): Zn 313 µg/g.
• Sample 49 — C. glomerata, “Putiki Point” (Waiheke Island): Zn 525 µg/g.
• Sample 50 — C. glomerata, “Matiatia Bay” (Waiheke Island): Zn 650 µg/g.
• Sample 51 — C. glomerata, “North Bay” (Ponui Island): Zn 350 µg/g.
• Sample 53 — C. glomerata, “Scullys Reef” (Ponui Island): Zn 475 µg/g.
• Sample 164 — P. novaezelandiae stomach (Tasman Bay): Cd 329 µg/g.
• Sample 167 — P. novaezelandiae stomach (Tasman Bay): Cd 257 µg/g.
• Sample 169 — P. novaezelandiae stomach (Tasman Bay): Cd 244 µg/g.
• Sample 170 — P. novaezelandiae stomach (Tasman Bay): Cd 217 µg/g.

Methods (brief)

Samples were collected from the intertidal zone by hand, by diving, or by wading, or from deeper-water species with one of two Fisheries Research Division vessels; specimens were returned frozen to the laboratory. A minimum of 14 average-sized individuals (unusually large or small individuals excluded) from each collection point were combined and homogenised in a Waring Blender to give a single average value representative of the local population. Aliquots of 10 g were dried 12 h at 100°C, then ashed 12 h at 450°C; metals in the ash were dissolved in 10 mL 3 M HNO₃ and analysed by atomic absorption spectrophotometry. Recovery experiments adding 2 µg of metal per millilitre of homogenate (three analyses per metal): Cd 100 ± 2%, Zn 99 ± 2%, Pb 93 ± 4%, Fe 84 ± 3%, Cu 100 ± 1%. Appendix values are not corrected for recovery. A hydrogen lamp was not used to compensate for scattering effects on lead, so lead values below 0.5 µg/g are flagged as questionable but included to indicate the probable existence of trace levels. Mercury was analysed separately by the Stux & Rothery (1971) cold-vapour method because the dry-ashing temperature volatilises mercury: 1.0 g (wet weight) sample was refluxed in 5 mL 16 M HNO₃ + 2 mL 36 M H₂SO₄ for 1 h and analysed by cold-vapour AAS; recovery of 1.0 µg/g added Hg was 100 ± 3%. All concentrations are reported on a wet-weight basis (µg per g wet); dry-weight figures are available from the authors but were not published because the paper is “orientated towards consumable levels of heavy metals.” The scallop Pecten novaezelandiae was analysed stomach-only by default (because the stomach uniquely accumulates cadmium per Brooks & Rumsby 1965), with gonad-only and adductor-only sub-samples for a small number of samples (Appendix footnotes *, †, ‡); other species were analysed as whole soft tissue. The map (Fig. 1) shows numbered sampling sites corresponding to Appendix 1 sample numbers.

Limitations

• No speciation: only total Hg by cold-vapour AAS (the dominant species in mollusc tissue is generally methylmercury, but the proportion is not characterised in this study); only total Pb (no organolead distinction); only total Cd (no protein-binding-form characterisation). Methylmercury, inorganic vs total arsenic, and chromium oxidation states are out of scope.
• Lead values below 0.5 µg/g are explicitly flagged by the authors as “questionable” because the AAS protocol did not include hydrogen-lamp compensation for scattering effects. These values are retained in the Appendix to indicate “probable existence of trace levels” but should not be treated as quantitatively reliable at that end of the range.
• Iron recovery was 84 ± 3% — significantly below the ~100% recovery achieved for the other metals — and Appendix iron values are not recovery-corrected. The authors’ decision to omit per-species iron averages is consistent with the lower analytical confidence for that metal in this dataset.
• Sample size per location is small (≥14 specimens per location, pooled before analysis), and most locations are sampled only once. The dataset gives one wet-weight number per location-by-species cell with no measure of within-location variance; site-to-site variation rather than within-site variation dominates the reported ranges.
• Geographic coverage is broad but is not a probability sample of the New Zealand coast; sampling was opportunistic and weighted toward edible species and accessible sites, with denser coverage in Wellington Harbour, Tasman Bay, and Foveaux Strait. The Foveaux Strait dataset (Samples 81-95, 146-147, 173-174, 186-188) was reported in a separate Nielsen (in press) paper on “Cadmium in New Zealand dredge oysters.”
• All concentrations are wet-weight on whole soft tissue (or specified sub-tissue for scallop and a handful of footnoted samples), with no separation of digestive gland, gill, mantle, or other organs except where explicitly noted. Whole-soft-tissue averages mask tissue-specific accumulation; the scallop stomach-vs-adductor-vs-gonad split is the only systematic tissue partitioning in the paper.
• Dating: the paper was submitted 19 July 1974 and revised 15 May 1975. Wellington P. canaliculus lead data are from January 1972; Tasman Bay scallop cadmium data are from July 1973. Several sites have undergone substantial industrial, urban, and regulatory change in the 50 years since (leaded petrol phased out in NZ in the 1990s; superphosphate cadmium content regulated more tightly; Tasman Bay scallop fishery closure history). The species-mean and site-level values document the 1972-1974 baseline; they should not be treated as current contamination levels.
• Mercury sub-sample size is small relative to the other metals (six species averages; Appendix shows mostly ”---” for Hg in non-subset samples), so the species-mean mercury figures are less robust than the Cd/Pb/Cu/Zn species means.
• Lead Sample 61 (C. glomerata, Omokoroa sewer outfall, Pb 21 µg/g) was excluded from the species average as an “abnormal sample” but retained in the species range. Treating it as an outlier rather than as a censored upper-bound is a choice; a meta-analytic pooling pass should decide explicitly how to treat such excluded point-source values.

Implications

• Occurrence and routing for shellfish ingredient page: the source contributes a 13-species, 199-site, six-metal national survey of New Zealand edible mollusc tissue (1972-1974), with whole-soft-tissue averages for Cd and Pb that include several species-mean cells above what later regulatory ceilings would impose. The dataset is routable to shellfish, bivalve-molluscs, molluscs, and seafood as a wet-weight occurrence record for the southwestern Pacific. The route to shellfish and seafood documents species-by-species variation that any future pooled-percentile analysis must respect (e.g., scallop adductor muscle vs scallop stomach are radically different Cd matrices; Crassostrea vs Perna vs Pecten have order-of-magnitude differences in Cu and Zn that pooling must keep separate).
• Certification context: this is a historical baseline record, not a basis for HMTc threshold-setting on contemporary shellfish products. Any pooled-percentile pass against modern bivalve regulatory ceilings (e.g., EU 2023/915 Cd in bivalves 1.0 mg/kg WW, Pb in bivalves 1.5 mg/kg WW; Codex Cd in bivalves 2.0 mg/kg WW; AU-NZ Food Standards Code Schedule 19) should treat the Nielsen & Nathan species means as a 1970s reference cell — the O. lutaria 3.9 µg/g Cd average and the Perna canaliculus 1.8 µg/g Pb average were measured before NZ phased out leaded petrol and before contemporary superphosphate-cadmium controls, and contemporary surveys (e.g., MPI shellfish biotoxin and contaminant monitoring) would be the operative source for current-state HMTc work.
• Courses: usable as a worked case in the seafood-and-aquaculture module for the species-specific physiological-accumulation theme: Crassostrea glomerata is the canonical Zn-and-Cu accumulator; Ostrea lutaria and Pecten novaezelandiae stomach are extreme Cd accumulators; Perna canaliculus is a Pb accumulator with a clear urban-runoff signal; Haliotis iris and Paphies species accumulate Cd and Pb to comparatively low levels and are arguably the “cleaner” edible bivalves on a 1970s New Zealand baseline. The Kawau Island copper-mine point-source dataset (380 → 305 → 165 µg/g Cu across the mine-to-40 m gradient) is a teachable example of how a defunct industrial site continues to load adjacent shellfish populations.
• App (if/when shipped): species- and site-specific bioaccumulation differences caution against treating “shellfish” as a single contamination class for likelihood-estimation purposes — at minimum, scallop-adductor vs scallop-stomach, oyster (Crassostrea, Ostrea) vs mussel (Perna, Mytilus, Modiolus, Aulacomya) vs clam/cockle (Paphies, Chione) vs gastropod (Haliotis), and locale (urban-proximate vs offshore) should be distinguishable. The 1970s baseline cannot be used directly for a contemporary likelihood estimator but is useful as the temporal-reference end of a multi-decade trend.
• Microbiome: out of scope. The source does not characterise mollusc gut microbiota or metal-microbe interactions.
• Cross-source signal: complements valencia2021-canacao-bay-oyster-pb-cd (Philippine cultured C. iredalei, 2016-2017), bjerregaard2023-mercury-mussels-denmark (Danish Mytilus, mercury), bao2024-sp-icp-ms-nps-mussels, bruno2024-metals-mussels-clams-faro-lake-sicily (Mediterranean bivalves), chou-uthe1993-cadmium-american-lobster-belledune (Canadian Atlantic point-source crustacean Cd), and any forthcoming MPI-NZ shellfish surveys as the 1970s baseline for southwestern Pacific edible-mollusc contamination. The dataset is unique in the corpus as a single-author, single-methodology, 13-species, 199-site sweep of one country’s coastal mollusc fauna.

Wiki pages this source may touch

• cadmium — 1970s NZ baseline; species-specific accumulation extremes (Ostrea lutaria 3.9 µg/g whole, Pecten novaezelandiae stomach 137 µg/g vs adductor 0.51 µg/g); Tasman Bay superphosphate-runoff hypothesis (Williams & David 1973).
• lead — 1970s NZ baseline; Perna canaliculus urban-proximity Pb accumulation (Wellington Harbour Fig. 2); leaded-paint and vehicle-exhaust pre-phase-out era.
• mercury — total-Hg subset (six species, all <0.5 µg/g averages, Mytilus edulis highest at 0.23); cold-vapour method.
• copper — Kawau Island defunct copper-mine point source (380 µg/g at the mine, gradient to 165 µg/g at 40 m); C. glomerata canonical Cu accumulator.
• zinc — C. glomerata extreme Zn accumulator (avg 337 µg/g, range 97-900); species-physiology contrast with O. lutaria (66 µg/g).
• iron — Fe up to 280 µg/g in Perna canaliculus; recovery 84% only.
• shellfish — 13-species national-survey wet-weight contamination panel.
• bivalve-molluscs — 12 of 13 species are bivalves; comprehensive bivalve panel.
• molluscs — includes Haliotis iris gastropod alongside the bivalves.
• seafood — southwestern Pacific occurrence record.
• shellfish — historical NZ shellfish Cd/Pb/Hg/Cu/Zn baseline.
• seafood — broader-context occurrence row.
• valencia2021-canacao-bay-oyster-pb-cd — companion cultured-oyster Pb/Cd dataset (Philippines, 2016-2017).
• bjerregaard2023-mercury-mussels-denmark — companion mussel mercury dataset (Denmark).
• bruno2024-metals-mussels-clams-faro-lake-sicily — companion Mediterranean bivalve multi-metal dataset.

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 10.1080/00288330.1975.9515582 not previously ingested; cite-key nielsen1975-nz-molluscs-heavy-metals not previously used; raw_handle KADC_heavy-metal-levels-in-new-zealand-molluscs not previously ingested).
• 2026-06-03 — Fresh-context audit subagent (verdict REVISE) flagged one ❌ definite error and one ⚠️ concern. (1) ❌ Opening narrative attributed Sample 61 to “Wellington-area sewer outfall”; verified against PDF p. 473 (text mentions sewer-outfall sample without naming the city) and Appendix p. 478 (Sample 61 listed under Bay of Plenty / “Omokoroa sewer outfall”) — the Wellington reference in the same paragraph attaches to the P. canaliculus 7.8 µg/g sample, not to Sample 61. Corrected to “Bay of Plenty sewer-outfall sample (Sample 61, Omokoroa)” in the opening narrative; the body and Selected-high-value-site-samples subsection already had the correct Omokoroa attribution. (2) ⚠️ Matrices abalone-paua flagged as non-standard; gastropod-mollusc (already in the matrices list) is the more general and convention-consistent matrix for Haliotis iris, so abalone-paua was removed. No ❌ numerical-fidelity issues in the Cd, Pb, Hg, Zn, or Cu species-average tables; Kawau Island Cu mine gradient (380 / 305 / 165), Tasman Bay scallop-stomach Cd extremes (Samples 164, 167, 169, 170), sample/site counts (203 samples, 199 sites, 13 species), and recovery values (Cd 100±2%, Zn 99±2%, Pb 93±4%, Fe 84±3%, Cu 100±1%, Hg 100±3%) all verified clean against the PDF.
• Full paper read directly from the PDF (16 pages including cover sheet, abstract through references and Appendix pages 477-481). Tables of species averages (Cd, Pb, Hg, Zn, Cu) in the body and the Appendix 1 sample-level table (Samples 1-203) transcribed; selected high-value site samples spot-cited from the Appendix to anchor the Cu mine-gradient and the scallop-stomach Cd extremes.
• Speciation flags: all reported as total elemental Cd, Pb, Cu, Zn, Fe by AAS, and total Hg (tHg) by cold-vapour AAS. Methylmercury was not measured; the methods explicitly use the Stux & Rothery 1971 cold-vapour procedure on HNO₃/H₂SO₄-digested samples, which captures total mercury without speciating MeHg vs iHg. The page uses tHg in the metals: array to reflect this.
• Units preserved as the source reports them: µg per g wet weight (= mg/kg WW = ppm WW) throughout, on whole soft tissue unless footnoted as gonad-only, stomach-only, adductor-only, or foot-only per Appendix legend. The dry-ashing step (12 h at 100°C, then 12 h at 450°C) is sample preparation, not a basis change. No silent conversion to ppb or to dry-weight.
• Method-vendor names retained per CLAUDE.md Part 12 Exception 2: Waring Blender (homogenisation). The paper does not name the AAS instrument vendor (uncommon for 1970s papers); cold-vapour mercury method follows Stux & Rothery 1971 (cited as a Varian Techtron technical-topics document, January 1971, Springvale, Australia).
• Brand-firewall: the source contains no consumer-brand contamination attributions. Geographic-place names (e.g., “Brampton Shoals,” “Kawau Island Copper Mine,” “Wairoa Bay,” “Omokoroa sewer outfall”) are place identifiers, not consumer brands, and are preserved as written. The Kawau Island copper mine is a defunct geographic feature, not a contemporary commercial operator; the steel/sewer outfall references are not named to any specific corporate entity in the source.
• Part 2 wiki/HMTc firewall: this page reports the Nielsen & Nathan 1972-1974 species means, the Appendix site-level samples, and the authors’ geological-volcanic and urban-runoff hypotheses. It does not propose, soften, or strengthen any HMTc shellfish threshold; the Implications section is restricted to noting routing destinations, the historical-baseline framing for any future HMTc pooling, and the dataset’s pre-leaded-petrol-phase-out temporal context.
• One contextual fact for downstream readers: the Appendix table cells in the PDF show some samples with a hand-written “YUCK” annotation on page 477 and crossed-out blocks on pages 478-480 (visible in the scanned PDF). These appear to be reader’s annotations on the archival copy, not authors’ marks; the underlying numeric data are otherwise legible and were transcribed directly from the printed values. None of the transcribed numbers in this wiki page depend on the annotated cells.
• The opening abstract sentence states “Cadmium, lead, copper, mercury, zinc and iron were analysed by atomic absorption spectrophotometry in 13 species of edible molluscs from 199 sites around New Zealand.” The Appendix Samples 1-203 reflect that some sites contributed multiple samples (e.g., separate cultured-mussel samples or replicate sub-tissue samples); the 199-site figure is from the abstract, and the 203-numbered-sample figure is from the Appendix table — both are stated in the source and not in conflict.

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.