La Pera et al. 2010 — Heavy metals and selenium in Sicilian virgin olive oils, by cultivar and crop year

La Pera and colleagues quantified Pb, Cd, Cu, Zn, and Se by derivative stripping chronopotentiometry in 49 virgin olive oils representing three Sicilian cultivars (Nocellara del Belice, Biancolilla, Cerasuola) across two crop years (2000 and 2001), then used canonical discriminant analysis to test whether metal and selenium fingerprints can classify oils by cultivar. Cadmium was below the limit of detection (<0.5 µg/kg) in every sample across both crop years; Pb means ranged from 26–50 µg/kg, well below the EC 466/2001 limit of 100 µg/kg for olive oil; Zn was the most abundant trace element measured (mean 162–335 µg/kg). The discriminant model classified 95.2% of 2000 samples and 85.7% of 2001 samples to their cultivar of origin, with Se and Cu contributing most to the discrimination — evidence that cultivar/geographic provenance fingerprinting from heavy-metal and selenium signatures is feasible for Sicilian PDO oils.

Key numbers

All concentrations are in µg/kg (= ppb) on a whole-oil basis as reported in Tables 18.6 and 18.8.

Method performance (Table 18.3): LODs — Cu 0.9, Pb 0.7, Cd 0.5, Zn 0.4, Se 0.6 µg/kg. LOQs — Cu 3.0, Pb 2.3, Cd 1.7, Zn 1.3 µg/kg for Zn, 2.0 µg/kg for Se. Recovery 93.5–97.0%. Precision (rsd %) 0.8–2.8. Linearity 0–1000 µg/kg (0–500 for Se).

2000 crop year, mean ± SD (max; min), µg/kg (Table 18.6):

• Nocellara del Belice (NB, n=8) — Pb 50.1 ± 10.1 (67.6; 36.3); Cd <0.5 (all samples); Cu 34.9 ± 10.3 (53.8; 21.6); Zn 189.9 ± 44.6 (250.0; 142.0); Se 53.5 ± 7.9 (68.0; 45.0)
• Biancolilla (BL, n=8) — Pb 31.0 ± 5.6 (39.4; 23.6); Cd <0.5; Cu 28.7 ± 12.1 (43.4; 14.0); Zn 161.7 ± 51.4 (214.1; 101.3); Se 14.9 ± 4.0 (19.6; 11.1)
• Cerasuola (CER, n=5) — Pb 40.7 ± 9.5 (56.0; 31.2); Cd <0.5; Cu 17.2 ± 5.3 (22.1; 8.2); Zn 190.4 ± 23.1 (217.5; 154.8); Se 34.7 ± 10.5 (49.1; 28.0)

2001 crop year, mean ± SD (max; min), µg/kg (Table 18.8):

• Nocellara del Belice (NB, n=10) — Pb 27.0 ± 7.5 (44.8; 17.8); Cd <0.5; Cu 21.2 ± 6.5 (Table prints “12.9; 34.7”, evidently transposed — see verification notes); Zn 168.9 ± 64.8 (271.7; 76.1); Se 60.2 ± 6.5 (70.1; 53.0)
• Biancolilla (BL, n=10) — Pb 26.2 ± 4.9 (Table prints “20.8; 34.6”, apparently transposed); Cd <0.5; Cu 16.6 ± 4.8 (Table prints “15.0; 24.7”, apparently transposed); Zn 174.6 ± 24.5 (Table prints “152.1; 220.0”, apparently transposed); Se 10.0 ± 2.6 (Table prints “5.3; 12.9”, apparently transposed)
• Cerasuola (CER, n=8) — Pb 27.8 ± 3.5 (34.4; 23.8); Cd <0.5; Cu 36.4 ± 15.3 (55.3; 17.4); Zn 335.0 ± 96.1 (485.9; 254.1); Se 36.9 ± 8.4 (49.1; 22.5)

Pooled summary across both crop years (derived from cell counts × means; this is a wiki-side derivation, not a table the authors print):

• Pb range across cultivar-by-year cells: 26.2 (BL 2001) – 50.1 (NB 2000) µg/kg
• Cd: <LOD (0.5 µg/kg) in all 49 samples
• Cu range: 16.6 (BL 2001) – 36.4 (CER 2001) µg/kg
• Zn range: 161.7 (BL 2000) – 335.0 (CER 2001) µg/kg
• Se range: 10.0 (BL 2001) – 60.2 (NB 2001) µg/kg

Regulatory context cited by authors (p. 155): EC Regulation 466/2001 set the maximum Pb in olive oil at 0.1 mg/kg (= 100 µg/kg). For Cd, Cu, and Zn no specific olive-oil legal limit existed under EC 466/2001 at time of writing. All Pb means in this dataset are well below the 100 µg/kg ceiling; the maximum single-sample Pb (NB 2000, 67.6 µg/kg) is also below it.

CDA classification performance (Tables 18.7, 18.9, and §18.5.3):

• 2000 crop alone (n=21): overall 95.2% correctly classified; NB 100% (8/8), BL 100% (8/8), CER 80% (4/5, 1 misclassified as BL). Wilks’ lambda significance test entered Se (F=25.85, λ=0.286) and Cu (F=6.33, λ=0.118) into the discriminant model; Zn and Pb were not retained.
• 2001 crop alone (n=28): overall 85.7%; NB 100% (10/10), BL 60% (6/10, 4 misclassified as NB), CER 100% (8/8). Discriminant model retained Zn (F=22.20, λ=0.367), Cu (F=14.23, λ=0.278), and Se (F=5.33, λ=0.180); Pb not retained.
• Both years pooled (n=49): 91.7% overall classification; NB and BL 100%, CER 70%. Se was the variable with the most significant cultivar-level variation, followed by Zn and Cu; Pb was the least discriminating.

Methods (brief)

Analytical technique. Derivative stripping chronopotentiometry (SCP), a stripping-voltammetry-class electrochemical method, on a three-electrode cell (mercury-film working electrode on glassy carbon, Ag/AgCl/3 M KCl reference, platinum wire auxiliary), with the dt/dE derivative transform applied during signal collection to eliminate non-faradaic current and adsorption interference. Four stripping mechanisms in this study (Table 18.2): chemical oxidation by Hg(II) (Chem. Ox.) for Cu, Cd, Pb, Zn; cathodic constant current (–30 µA CCC) for Se(IV) (cathodic stripping mode); accumulation 120 s at –1200 mV (Cu/Cd/Pb/Zn) or –150 mV (Se); discharge potentials –260 (Cu), –430 (Pb), –640 (Cd), –950 (Zn), –580 (Se) mV. Acid extraction: 3.0 g oil aliquot, 2.0 mL 35% H₂O₂, 10.0 mL 36% HCl in a Teflon beaker, magnetic stirring at 90 °C for ~30 min, then 5 min centrifugation at 4800 rpm; acid phase taken, organic layer re-extracted with 8.0 mL boiling water, combined; passed through Supelclean ENVI-Carb SPE (0.5 g, 6 mL) to remove organic residues; acid filtrate analyzed by SCP.

Speciation flag. Pb, Cd, Cu, Zn reported as total metal in the acid-digested oil matrix; no speciation. Se reported as Se(IV) — the cathodic-stripping species after digestion in HCl/H₂O₂, which oxidizes Se(–II) and Se(0) to Se(IV). Total Se in oil is therefore captured to the extent the digestion fully oxidizes organoselenium compounds; the paper does not characterize this completeness explicitly.

Sample design. 49 virgin olive oil samples (Table 18.4): 18 NB (Valle del Belice, 8 from 2000 + 10 from 2001), 18 BL (Val di Mazara, 8 + 10), 13 CER (Valli Trapanesi, 5 + 8). Sicilian PDO/PGI growing zones. No sample-by-sample data are tabulated; only by-cultivar means, max, min, and SD.

Statistical analysis. Stepwise forward canonical discriminant analysis (CDA) on metal-concentration vectors per sample, with cultivar as grouping variable; ANOVA at α=0.05 for univariate variance; Wilks’ lambda F-test for variable selection. SPSS 12.0 (2003) and STATISTICA 6.0 (2000).

Limitations. (1) The 2001 cultivar-by-cultivar max/min values printed in Table 18.8 have transposed pairs in several cells (see Key numbers); means and SDs appear internally consistent and are taken as authoritative. (2) Cadmium was below the method LOD of 0.5 µg/kg in all 49 samples — useful as a ceiling statement (Cd in Sicilian virgin olive oil from these three cultivars is < 0.5 µg/kg in 2000–2001) but no positive Cd characterization is possible from this dataset. (3) No certified reference material specifically for olive oil is cited for method validation; accuracy for Cd, Cu, Pb, Zn was validated via recovery tests against CRMs cited in Dugo et al. 2004b, and Se accuracy via spike/recovery on a selenium-free olive-oil sample (Dugo et al. 2004a). (4) Sample selection (which farms within each PDO zone) is not described in detail.

Implications

Certification. This dataset is one of the cleanest by-cultivar olive-oil heavy-metal characterizations available for Sicilian virgin oils. For HMTc purposes: (a) Cd in Sicilian virgin olive oil from these three cultivars is consistently <0.5 µg/kg — useful as a literature-baseline P97 ceiling input where pooled with sibling olive-oil sources; (b) Pb means by cultivar/year span 26–50 µg/kg, with single-sample maxima up to 67.6 µg/kg, all under the EC 466/2001 100 µg/kg ceiling but materially above the 10 µg/kg IOC trade specification that some other olive-oil sources cite. The gap between 26–50 µg/kg observed and 10 µg/kg IOC-aspirational deserves explicit treatment in olive-oil threshold-setting; this paper supports a literature-baseline > 10 µg/kg for Pb in Sicilian virgin olive oil at the population level.

Courses. Useful teaching case for (a) the difference between by-cultivar fingerprinting (the paper’s primary purpose) and by-population contamination characterization (HMI’s primary purpose) — the same dataset serves both, but only the second is contamination-relevant; and (b) the limits-of-detection problem: Cd at <0.5 µg/kg returns “negative” Cd information that is still useful as a ceiling, but cannot be averaged or P-percentiled meaningfully.

App. Sicilian PDO virgin olive oil priors: Pb ~ 26–50 µg/kg, Cd <0.5 µg/kg, Cu 16–36 µg/kg, Zn 162–335 µg/kg, Se 10–60 µg/kg. These are means by cultivar/crop-year cell, not population-weighted. Cerasuola from the 2001 crop is the highest-Zn cell; Nocellara del Belice from 2001 is the highest-Se cell. Cd ceiling holds across all 49 samples.

Speciation caution. Lead, Cd, Cu, Zn are total metal in acid-digested oil. Se is reported as Se(IV) after HCl/H₂O₂ digestion, which captures inorganic Se and the inorganic fraction of organoselenium; the paper does not characterize total-Se completeness independently. Inorganic-vs-organic Se in olive oil is not addressed.

Provenance discriminator. The CDA result — that Sicilian olive cultivars can be classified at 86–95% accuracy from a five-element fingerprint dominated by Se, Cu, Zn — is methodologically interesting for traceability/anti-fraud applications but does not bear directly on contamination thresholds. It is logged here so the source can be cited from olive-oil pages on the question of whether cultivar/origin meaningfully predicts contamination, not just botanical identity.

Wiki pages this source may touch

• olive-oil
• olives
• olive-oil
• lead
• cadmium
• copper
• zinc
• eu-466-2001-contaminants-superseded

Verification notes

• 2026-06-01 fresh ingest (Claude, MFK June 1 batch). Cite-key la-pera2010-sicily-olive-oil-metals-selenium. No prior wiki source page matched on DOI/handle/cite-key — NEW path.
• Source-internal data-integrity note (Table 18.8). Multiple cells in Table 18.8 print max/min pairs in apparently transposed order — e.g., BL 2001 Pb prints “max 20.8; min 34.6” (mathematically max < min). The likely typographical pattern is consistent across BL 2001 (Pb, Cu, Zn, Se all transposed) and partially in NB 2001 (Cu only). Cell means and SDs are mathematically self-consistent and are used as authoritative; the printed max/min values are reproduced verbatim with the “Table prints” qualifier so future readers see what the source actually shows. This is the obvious mirroring pattern referenced in the v2 skill’s stop-condition list — paper-internal contradiction is logged but does not block ingest.
• 2001 sample count discrepancy. §18.5.2 reads “24 samples of Sicilian virgin olive oils produced from three different cultivars in 2000 crop year” — the year is a typo (should read 2001), and the count (24) disagrees with Table 18.4 which gives 2001 totals of 10 + 10 + 8 = 28. Table 18.4 is taken as authoritative for sample counts.
• DOI. No DOI is printed on the chapter pages read; book chapter DOIs in this Elsevier volume typically follow the form 10.1016/B978-0-12-374420-3.00018-X but I do not have authoritative confirmation. Frontmatter doi: left null with no_doi_assigned: true marker per the wiki’s DOI-fallback convention; access_url set to the publisher’s parent-volume page at sciencedirect.com (ISBN 9780123744203). ISBN 978-0-12-374420-3 and chapter number 18 are also captured in the publication: field.
• License. Copyright © 2010 Elsevier Inc., “All rights of reproduction in any form reserved.” Not open access; license: set to “not confirmed” pending licensing-database check.
• Metals frontmatter. Includes Se because the paper measures it; no metals/selenium wiki page exists in the current taxonomy snapshot, so no [[metals/selenium]] wikilink appears in the source-touches section. Routing-audit may treat this as soft-unmatched; the analyte should still appear in the frontmatter for correct routing coverage.
• Brand firewall. No brand names appear in source; cultivar names (Nocellara del Belice, Biancolilla, Cerasuola) and PDO zones (Valle del Belice, Val di Mazara, Valli Trapanesi) are botanical/geographic, not brand identifiers.
• Regulation page reference. EC 466/2001 is cited by the paper as setting Pb at 0.1 mg/kg for olive oil; the wiki regulation slug eu-466-2001-contaminants-superseded is the correct (now-superseded) referent. The current EU regulation on Pb in olive oil is captured separately and is outside this paper’s scope.

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.