Dearing et al. 2025 — Heavy metals in Hawke’s Bay vegetables one year after Cyclone Gabrielle, with organic vs. non-organic and flooded vs. non-flooded comparisons
Dearing and colleagues sampled 736 commercially grown vegetables (combined into 153 composite representative samples) from 14 market gardens at 10 growing sites in Hawke’s Bay, New Zealand, in February 2024, one year after Severe Tropical Cyclone Gabrielle caused widespread flooding in February 2023. The study compared organic vs. non-organic and flooded vs. non-flooded growing as factors. Cadmium and nickel were significantly lower in organic vegetables and in vegetables grown on flood-affected land; mercury was below LOD in all samples; three samples exceeded the Codex 0.1 mg/kg FW lead limit and one Brassica exceeded the 0.05 mg/kg FW cadmium limit.
Key numbers
- 736 individual vegetables combined into 153 representative composite samples (~200 g each), each composite drawn from one vegetable type from one market (Methods).
- Sampling from 14 market gardens at 10 growing sites within 50 km of Napier and Hastings; 4 organic market gardens at 6 growing sites; sampling month February 2024 (Methods).
- All metals reported on fresh-weight (FW) basis (Methods).
- Table 2 — descriptive statistics (mg/kg FW, all n=153 unless excluded):
- Cadmium: mean 0.011, SD 0.014, median 0.005 (including all 153); excluding samples below LOD (n=131): mean 0.013, SD 0.014, median 0.007, max 0.093.
- Lead: mean 0.014, SD 0.058, median 0.005 (including all); excluding LOD (n=12): mean 0.123, SD 0.181, median 0.047, max 0.61.
- Arsenic: mean 0.014, SD 0.02, median 0.01 (including all); excluding LOD (n=2): mean 0.18, SD 0.085, median 0.18, max 0.24.
- Nickel: mean 0.067, SD 0.139, median 0.035 (including all); excluding LOD (n=103): mean 0.097, SD 0.163, median 0.055, max 1.5.
- Chromium: mean 0.02, SD 0.072, median 0.01 (including all); excluding LOD (n=4): mean 0.33, SD 0.365, median 0.208, max 0.84.
- Thallium: mean 0.005, SD 0.009, median 0.003 (including all); excluding LOD (n=10): mean 0.034, SD 0.04, median 0.037, max 0.058.
- Mercury: below LOD (0.01 mg/kg FW) in all 153 samples and excluded from statistical analysis.
- Three lead-exceeding samples (two Lactuca sativa and one Petroselinum crispum) above Codex 0.1 mg/kg FW; one cadmium-exceeding sample (Brassica rapa) above 0.05 mg/kg FW.
- Top reported individual values: lead 0.61 mg/kg FW and 0.37 mg/kg FW (both in lettuce composites).
- Table 3 — Cd and Ni by genus (median; max, mg/kg FW): Allium 0.005; 0.029 / 0.010; 0.170. Apium 0.016; 0.019 / 0.027; 0.030. Beta 0.015; 0.060 / 0.021; 0.340. Brassica 0.006; 0.093 / 0.035; 0.270. Cucumis 0.001; 0.004 / 0.067; 0.094. Cucurbita 0.003; 0.005 / 0.057; 1.500. Foeniculum 0.003; 0.004 / 0.041; 0.072. Ipomoea 0.019; 0.019 / 0.040; 0.040. Lactuca 0.025; 0.074 / 0.032; 0.610. Lagenaria 0.003; 0.003 / 0.068; 0.082. Luffa 0.009; 0.015 / 0.093; 0.150. Momordica 5.0×10⁻⁴; 5.0×10⁻⁴ / 0.057; 0.057. Petroselinum 0.013; 0.028 / 0.155; 0.370. Phaseolus 7.5×10⁻⁴; 0.003 / 0.12; 0.160. Raphanus 0.011; 0.027 / 0.005; 0.005. Solanum 0.003; 0.013 / 0.005; 0.070. Zea 0.001; 0.003 / 0.054; 0.110.
- ANOVA (Table 4) — Cadmium ~ organic (F=9.198, p=0.003, η²p=0.058, 95% CI 0.007–0.145) and ~ flooded (F=4.817, p=0.030, η²p=0.031, 95% CI 0.000–0.104); interaction not significant (p=0.270).
- ANOVA (Table 5) — Nickel ~ organic (F=21.038, p<0.001, η²p=0.124, 95% CI 0.042–0.227) and ~ flooded (F=5.164, p=0.024, η²p=0.033, 95% CI 0.000–0.108); interaction p=0.010.
- Sensitivity sub-analysis (Table 6, n=37 across broccoli, cabbage, and lettuce): Cadmium not-organic flooded mean 0.011 SD 0.012, not-organic not-flooded 0.026/0.023, organic flooded 0.005/0.008, organic not-flooded 0.009/0.007. Nickel not-organic flooded 0.050/0.049, not-organic not-flooded 0.202/0.196, organic flooded 0.019/0.014, organic not-flooded 0.075/0.10.
Methods (brief)
736 vegetables were collected from 14 market gardens at 10 growing sites in Hawke’s Bay during February 2024 and combined into 153 composite representative samples (~200 g each) per vegetable type per market. Edible portions (with onion outer skin and root-vegetable skin removed to replicate normal food preparation) were washed in fresh running tap water, weighed for fresh weight, oven-dried at 60 °C to constant weight, ground, and digested by aqua regia, then diluted to 2% HNO3 / 1% HCl prior to ICP-MS analysis in an ISO 17025-accredited laboratory. Mercury LOD 0.01 mg/kg FW; mercury was below LOD in all samples and excluded from statistical analysis. All metals are reported as total elemental concentrations on fresh-weight basis; no speciation was performed. Statistical analyses computed two sets of descriptive statistics (excluding LOD values and using half-LOD for LOD values), with log10 transformation prior to ANOVA where required for normality. Growing-site flood-impact status was verified using digital flood maps from Toitū Te Whenua Land Information New Zealand. Sub-sample composite design masks within-vegetable variability; authors note that representative sample maxima still exceeded Codex limits in four cases.
Implications
- Certification (HMTc): Adds 153 New Zealand market-garden composite-sample occurrence rows for Cd, Pb, tAs, Ni, Cr, and Tl across 17 vegetable genera on fresh-weight basis. Mercury non-detection across the corpus is itself useful evidence for NZ-market vegetables. Organic vs. non-organic and flooded vs. non-flooded sub-analyses are relevant to sourcing-lever discussions on ingredient pages.
- Courses: Suitable case study on the effect of farming practice (organic vs. conventional) and post-disaster flooding on heavy-metal accumulation in commercial vegetables, for QA and supply-chain audiences.
- App: Contributes occurrence data points to lettuce (Pb-exceeding samples), parsley (Pb-exceeding sample), cabbage / broccoli / cauliflower (Brassica Cd exceedance), and broad vegetable genera Cd and Ni medians.
Wiki pages this source may touch
- vegetables
- leafy-vegetables
- root-vegetables
- lettuce
- parsley
- cabbage
- broccoli
- cauliflower
- onions
- potatoes
- tomato
- corn
- green-beans
- celery
- beet
- butternut-squash
- cadmium
- lead
- nickel
- arsenic-total
- chromium
- thallium
- mercury-total
- codex-cadmium-mls
- eu2023-contaminants-maximum-levels
Verification notes
- 2026-05-20 — Re-audited against PDF (raw/Manual Fetch Kimi /02_Vegetables_and_Vegetable_Products/02_Vegetables_and_Vegetable_Products/Assessment of Heavy Metals in Organic and Non-Organic Vegetables in Hawkes Bay, New Zealand.pdf). All 153-sample descriptive statistics in Table 2 (Cd, Pb, As, Ni, Cr, Tl, Hg including/excluding LOD), all Table 3 genus medians/maxes for Cd and Ni across 17 genera, ANOVA outputs from Tables 4 and 5, and Table 6 sensitivity sub-analysis (n=37) values verified line-by-line — all match. Exceedance counts verified (three lead exceeding Codex 0.1 mg/kg FW: two Lactuca sativa + one Petroselinum crispum; one Brassica rapa exceeding 0.05 mg/kg FW Cd). Methods description (ICP-MS, ISO-17025 lab, aqua regia digestion, 2% HNO3/1% HCl dilution, oven-dry at 60 °C, fresh-weight basis, half-LOD substitution + LOD-excluded duplicate stats) matches PDF. Speciation conventions correct (tAs not iAs, tHg not MeHg, Cr not Cr-VI — paper used total elemental ICP-MS without speciation).
- Added root-vegetables to ingredients array and to wiki-pages list — root-vegetable was already in matrices but the parent ingredient slug was missing.
- products: [] is intentional. This is raw fresh produce; no packaged-product slug applies. The routing-malformed advisory note (“Source is missing optional routing-input fields”) is expected and does not indicate a defect.
- Genera mentioned in PDF whose ingredient slugs do not yet exist (not added to ingredients array, not stubbed here): kale, bok-choi, red-cabbage (Brassica sub-types), garlic, leek, red-onion, spring-onion (Allium sub-types), chard, silver-beet (Beta sub-types), beetroot (Beta — beet slug used as parent), chinese-winter-melon, courgette (Cucumis), grey-pumpkin (Cucurbita — butternut-squash slug used as parent), fennel (Foeniculum), chinese-water-spinach (Ipomoea), bottle-gourd (Lagenaria), luffa-gourd (Luffa), bitter-melon (Momordica), daikon/diakon (Raphanus), cherry-tomatoes (Solanum — tomato slug used as parent), sweet-corn (Zea — corn slug used as parent). Auto-stub freq-counter will catch the freq≥2 cases across the corpus.
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.