Bora et al. 2015 — Heavy metals in the soil-grapevine-wine system in NW Romania

Bora and colleagues measured concentrations of ten elements (Na, Ca, Mg, Fe, Cu, Zn, Pb, Cd, Ni, Co) by flame atomic absorption spectrophotometry (FAAS) across the soil–grapevine–wine system of three Vitis vinifera cultivars (Feteasca albă, Feteasca regală, Riesling italian) grown in the Turulung vineyard (Satu Mare County, NW Romania), 57 km from the Baia Mare metallurgical complex. The vineyard soil exceeded the Romanian maximum limit allowed (MLA) for copper (mean 479.64 mg/kg vs 20 mg/kg MLA), attributed by the authors to long-running Cu-based fungicide treatments and regional smelter emissions; all other soil analytes (Pb, Cd, Zn, Ni, Co) were below their MLAs. In the resulting micro-vinified young wines, all measured heavy metals were below the International Organisation of Vine and Wine (OIV) ceilings: Cu 0.23 mg/L, Zn 0.46 mg/L, Pb 0.09 mg/L, Ni 0.04 mg/L; Cd was under the limit of detection. Transfer-factor analysis showed Cd and Pb preferentially accumulating in leaves rather than in must, consistent with vine-internal sequestration of toxic metals away from the berries.

Key numbers

All concentrations are mean ± standard deviation in mg/kg dry weight for solid matrices and mg/L for liquids unless noted. Replication structure: n = 3 samples per matrix per cultivar. Soil sampled at four depths (0–20, 20–40, 40–60, 60–80 cm) over one 4-ha plot. Analytical method: FAAS (Perkin Elmer AAnalyst 800). Per-element LODs from Table 6 (mg/L): Pb 0.051, Cd 0.028, Cu 0.017, Zn 0.018, Ni 0.020, Co 0.120, Ca 0.092, Mg 0.190, Fe 0.110, Na 0.012. MLA = Romanian regulatory maximum limit allowed (Order of the Ministry of Agriculture No. 756/3 November 1997).

Soil, 0–80 cm pool (Table 1; mg/kg dry):

• Cu: 479.64 ± 53.97; range 433.69–535.58 — exceeds MLA 20 mg/kg (the only soil analyte above its MLA)
• Zn: 69.44 ± 4.02; range 54.58–78.25 — below MLA 100
• Pb: 14.77 ± 0.74; range 13.63–16.27 — below MLA 20
• Cd: 0.45 ± 0.10; range 0.30–0.56 — below MLA 1
• Ni: 16.28 ± 2.01; range 12.80–19.23 — below MLA 20
• Co: 9.75 ± 1.47; range 7.49–11.31 — below MLA 15

Soil 20–80 cm root-zone average (Table 1; the authors note vine roots concentrate in this interval per Hooda 2010): Cu 494.95 ± 58.35, Zn 66.51 ± 3.86, Pb 14.27 ± 0.27, Cd 0.41 ± 0.03, Ni 15.30 ± 2.36, Co 9.69 ± 1.39 mg/kg.

Canes, 3-cultivar average (Table 2; mg/kg dry):

• Cu: 55.02 ± 2.54
• Zn: 13.97 ± 1.30
• Pb: 2.02 ± 0.19
• Cd: 0.13 ± 0.02
• Ni: 9.87 ± 0.44
• Co: 1.79 ± 0.28

Leaves, 3-cultivar average (Table 3; mg/kg dry):

• Cu: 49.16 ± 3.40
• Zn: 25.20 ± 1.37
• Pb: 4.25 ± 0.03
• Cd: 0.77 ± 0.03
• Ni: 9.47 ± 0.75
• Co: 0.47 ± 0.07

Wine, 3-cultivar mean (Table 4 “This study” row; mg/L):

• Cu: 0.23
• Zn: 0.46
• Pb: 0.09
• Ni: 0.04
• Cd: under limit of detection (ULD)

Table 4 places these Turulung wine values alongside other European studies (Spain Cu 0.30 mg/L, Slovenia Cu 0.12 mg/L, Romania/Moldova Cu 0.602 mg/L per Geana et al., Ukraine Cu 0.48 mg/L, Greece Cu 0.2–0.6 mg/L) and against the OIV recommended ceilings (Cu 1.0, Zn 5, Pb 0.15, Cd 0.01 mg/L). The Turulung wines were below every OIV ceiling, including the 0.15 mg/L Pb ceiling.

Grape juice (must), per cultivar (text, p. 7; mg/L):

• Cu (Feteasca albă): must 9.32 → wine 0.23
• Cu (Feteasca regală): must 8.77 → wine 0.32
• Cu (Riesling italian): must 10.22 → wine 0.15
• Zn in white must: 5.02 (Feteasca regală) to 7.93 (Feteasca albă)
• Cd in must: under limit of detection in all three cultivars
• Cu, Zn, Pb and Ni all decreased between must and wine, which the authors attribute to insoluble metal-fraction sedimentation with yeast lees during fermentation.

Transfer factors (Table 5; dimensionless; cultivar average; ranking notes preserved from source):

• TF_cs (canes ÷ soil): Cu 0.111, Zn 0.215, Pb 0.142, Cd 0.325, Ni 0.645 — order Ni > Cd > Zn > Pb > Cu
• TF_lc (leaves ÷ canes): Cu 0.933, Zn 1.797, Pb 2.144, Cd 6.334, Ni 0.968 — order Cd > Pb > Zn > Ni > Cu
• TF_mc (must ÷ canes): Cu 0.180, Zn 0.438, Pb 0.128, Cd 0.000, Ni 0.019 — order Zn > Cu > Pb > Ni
• TF_wm (wine ÷ must): Cu 0.025, Zn 0.078, Pb 0.330, Cd 0.000, Ni 0.217 — order Pb > Ni > Zn > Cu

Relative standard deviations on the cultivar-average transfer factors (Table 5; n = 3) range from 10.7% (TF_lc Pb) to 86.97% (TF_wm Pb). The high leaves-to-canes transfer factors for Cd (6.334) and Pb (2.144) are the highest in the system and are interpreted by the authors as a vine-internal avoidance strategy that diverts toxic metals into leaves rather than into berries.

Methods (brief)

Study area: Turulung vineyard (47°56′N, 23°5′E; 133 m elevation; 687 mm annual precipitation; 10.3 °C annual mean temperature), Satu Mare county, NW Romania, 57 km northwest of Baia Mare. Soil type: Preluvosoil (EL) from Luvisols (LUV) class; clay-loam texture, pH 6.0–6.12, 1.3–3.5% humus. Sampling window: August–October 2013. Soil sampled at four depths (0–20, 20–40, 40–60, 60–80 cm) per ISO 11464/1994 with a handle steel sampler, 3 samples per depth. Canes (2–4 pieces, 25 cm) and leaves (2–4 per vine) drawn from 10–12 representative vines per cultivar, washed in distilled water, oven-dried at 80 °C to constant weight, ground, sieved to 2 mm, then mineralised at 550 °C for 8 h. Grapes: 5 kg per cultivar from 10–12 vines, cold-pressed to must; young wines (3 samples per cultivar) produced by laboratory micro-vinification (November 2013) and analysed without further pretreatment.

Digestion: 0.2–0.5 g of milled solid material digested in 12 mL aqua regia (9 mL HCl + 3 mL HNO3) using a Berghof MWS-2 microwave (180 °C then 100 °C, two steps), filtered through 0.45 µm, brought to 100 mL with ultrapure water per ISO 11466/1999. Vegetable ash was redissolved in 2 mol/L HNO3 to 100 mL.

Quantification: flame atomic absorption spectrophotometry (FAAS) on a Perkin Elmer AAnalyst 800 (Shelton, USA). Air–acetylene flame at 2300 °C for all ten elements. Background correction: deuterium for Mg, Fe, Zn, Cu, Ni, Pb and Cd; halogen for Co; none for Ca and Na. Wavelength, slit, correlation coefficient, LOD and LOQ are tabulated per element in Table 6. Calibration: Merck stock standard solutions (1000 mg/L) and Barnstead Easypure RoDi D13321 high-purity water; glassware soaked 24 h in 10% v/v HNO3. QC: blanks and triplicate samples (n = 3) per run; variation coefficients reported under 10%. Statistics: ANOVA with Duncan multiple range tests (PoliFact 2010 ANOVA and Duncan’s Test PC Program); significance at p < 0.05.

Limitations explicitly visible in the source: (i) single vineyard, single harvest year (2013); (ii) micro-vinification rather than commercial production, which the authors flag as a likely under-estimate of commercial-wine metal content; (iii) no certified reference material (CRM) recovery is reported; (iv) arsenic and mercury were not measured; (v) chromium was not measured, so Cr-VI speciation is absent; (vi) replication is n = 3 per cultivar per matrix, with transfer-factor RSDs up to 87% (Table 5 TF_wm Pb); (vii) Table 5 prints the TF_cs Pb value for Feteasca albă as “0. 1 60” with embedded spaces — a typesetting artefact for 0.160.

Evidence Fitness

This source provides occurrence values for wine, grape must, grape leaves, canes and vineyard soil at one 4-ha vineyard in NW Romania under documented regional Cu pressure (Baia Mare metallurgical complex, 57 km) and explicit historical Cu-fungicide use. The cultivar set (n = 3) is small and is constrained to one harvest year (2013); the wines are laboratory micro-vinifications rather than commercial product. It can support discovery, geographic context for NW Romanian terroir under known Cu loading, a methods anchor (FAAS with documented LOD/LOQ and aqua-regia microwave digestion), and an illustration of vine cane→leaf accumulation patterns for Pb and Cd. It is not large enough or structured enough to support pooled-percentile work on the fermented-beverages-non-tea-based row. Reported public evidence label: Context only.

Implications

• Certification: contributes wine-stage occurrence points (Cu 0.23, Zn 0.46, Pb 0.09, Ni 0.04 mg/L; Cd ULD) and vineyard-input context to fermented-beverages-non-tea-based for Cu, Zn, Pb, Cd, Ni and Co. The wine values sit below OIV ceilings and below the higher Romanian-region values reported by Geana et al. (Cu 0.602 mg/L). Sample size is too small to drive pooled-percentile selection but anchors a NW Romanian micro-vinification data point.
• Courses: useful as a teaching case for vine-internal partitioning of toxic metals — TF_lc of 6.334 for Cd and 2.144 for Pb demonstrates that grapevine routes Pb and Cd into leaves rather than into berries, with downstream implications for grape-leaf-as-food uses (dolma, sarma) but not for the wine product itself. Also useful as a worked example in which extreme soil Cu (479.64 mg/kg, 24× MLA) does not translate to wine Cu exceeding OIV ceilings, illustrating the role of fermentation-stage sedimentation in metal removal.
• App: contributes Cu, Zn, Pb, Ni occurrence points for Romanian wine; Cu, Zn, Pb, Cd, Ni, Co occurrence points for grape leaves, grape canes, grape must and vineyard soil under documented regional smelter influence.

Provenance notes

Open-access article published 12 April 2015 in Chemistry Central Journal (BioMed Central) under the Creative Commons Attribution License 4.0; the CC Public Domain Dedication 1.0 applies to the data per the licensing statement on page 1. Accessed via the Manual Fetch Discovery autopilot.

Evidence tier set to B rather than A. The journal is peer-reviewed and the methods section is unusually transparent (LOD, LOQ, wavelength, slit, correlation coefficient and background-correction mode per element are tabulated in Table 6). What disqualifies it from A on the HMI scale is the absence of a certified reference material recovery, the single-vineyard / single-harvest design, the n = 3 replication structure that produces transfer-factor RSDs up to 87%, and the use of laboratory micro-vinification rather than commercial production for the wine endpoint. The methodological transparency on the analytical side is strong; the experimental design is descriptive rather than confirmatory.

Wiki pages this source may touch

• lead
• cadmium
• nickel
• cobalt
• copper
• zinc
• grapes
• grape-juice
• fermented-beverages-non-tea-based

Verification notes

No brand names appear in the paper or on this page. Cultivar names (Feteasca albă, Feteasca regală, Riesling italian) are Vitis vinifera grapevine varieties, not commercial wine brands; Part 12 brand-firewall does not apply. The regional geographic indication “Sătmar Hills” identifies the production area, not a brand. Instrument and reagent references (Perkin Elmer AAnalyst 800; Berghof MWS-2 microwave; Merck standard solutions; Barnstead Easypure RoDi D13321) are reproduced under the methods-vendor exception (Part 12, Exception 2).

The metals frontmatter list contains only the six measured trace/heavy metals (Pb, Cd, Cu, Zn, Ni, Co); the four measured macroelements (Na, Ca, Mg, Fe) are reported in the body where relevant to soil context but are not in the frontmatter metals list, in line with HMI scope on toxic metals plus Cu/Zn/Ni as toxic-borderline elements. Arsenic, mercury and chromium were not measured, so iAs, tAs, MeHg, tHg, Cr and Cr-VI are absent.

The matrices grape-leaves and grape-canes are new bare-string additions to the matrices vocabulary, consistent with the existing pattern (olive-leaves, rapeseed-leaves, cacao-leaf). wine, grape-juice and soil are already in use across wiki/sources/.

Wine values in Table 4 (“This study” row) are the 3-cultivar mean. Per-cultivar wine Cu values are quoted in the text (Feteasca albă 0.23, Feteasca regală 0.32, Riesling italian 0.15 mg/L), which average to 0.233 mg/L, consistent with the 0.23 mg/L printed in Table 4. Figure 1 reports per-cultivar wine bar heights for Cd, Cu, Ni, Pb and Zn but the axis values are read from a bar chart and are not tabulated; this page does not reproduce figure-derived per-cultivar wine concentrations to avoid attributing precision the source does not print.

The Table 5 TF_cs Pb value for Feteasca albă is printed in the source as “0. 1 60” with embedded spaces, a typesetting artefact for 0.160. The page reports the 3-cultivar average TF_cs Pb of 0.142 and does not reproduce the typographic artefact.

Table 5 prints the TF_wm right-margin ranking annotation as “Pb > Ni > Cu > Zn”, which contradicts its own averages (Cu 0.025 < Zn 0.078). The arithmetically correct descending order is Pb > Ni > Zn > Cu, which is what this page reports. The source annotation appears to be a typo; the underlying numeric averages are not in dispute.

Tables 1, 2 and 3 in the source print “Average” rows whose ± values are implausibly tight relative to the per-cultivar spreads in the same tables (e.g., Table 3 leaves Pb Average 4.25 ± 0.03 vs per-cultivar values 4.93 / 3.12 / 4.69, whose cross-cultivar SD is ≈ 0.99; Table 3 leaves Cd Average 0.77 ± 0.03 vs per-cultivar 0.58 / 0.79 / 0.93, cross-cultivar SD ≈ 0.18). The Average-row ± appears to be the within-replicate (n = 3) analytical SD averaged across cultivars rather than the cross-cultivar SD. This page reproduces what the source prints; downstream users should not treat the Average-row SDs as cross-cultivar variability.

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.