Findings and scope
The study measured lead, total arsenic, total chromium, nickel, zinc and copper in 56 potato tuber samples and 56 paired topsoil samples from 15 towns in Weining county, Guizhou province, a karst area with high geochemical background and a history of artisanal zinc smelting (PDF p.2). Mean tuber lead was 0.027 mg/kg fresh weight (maximum 0.095), total chromium 0.072 mg/kg (maximum 0.659) and nickel 0.055 mg/kg (maximum 0.202); total arsenic was below the 0.01 mg/kg detection limit in every tuber (Table 3, PDF p.7). Soils carried a mean of 59.3 mg/kg lead and 25.8 mg/kg total arsenic, and the authors report that 39 of 56 soils exceeded their 20 mg/kg arsenic limit and 35 exceeded their 100 mg/kg chromium limit (Table 2, PDF p.5). Tuber metals were weakly related to soil metals; the only significant soil-to-tuber correlation was soil arsenic with tuber lead (r = 0.301, p < 0.05), and soil pH was negatively correlated with tuber nickel and chromium (Table 5, PDF p.8). Cadmium was not measured in this study; the paper quotes the authors’ earlier Weining work for cadmium (soil mean 2.60 mg/kg, tuber mean 0.083 mg/kg fresh weight), which is secondhand here (PDF p.5 and p.6). The study analysed raw tubers only and contains no data on potato starch.
Key numbers
Tuber values as printed in Table 3 (PDF p.7), mg/kg fresh weight, n = 56. Soil values from Table 2 (PDF p.5), mg/kg, air-dried soil, n = 56.
| Metal | Tuber mean ± SD | Tuber median | Tuber range | Soil mean ± SD | Soil median | Soil range |
|---|---|---|---|---|---|---|
| lead (Pb) | 0.027 ± 0.018 | 0.024 | 0.010 to 0.095 | 59.3 ± 20.9 | 52.7 | 20.0 to 115 |
| total arsenic (tAs) | < 0.010 | < 0.010 | all < 0.010 | 25.8 ± 10.1 | 23.5 | 9.74 to 62.8 |
| chromium (Cr) | 0.072 ± 0.153 | 0.023 | 0.010 to 0.659 | 119 ± 41.2 | 108 | 62.1 to 292 |
| nickel (Ni) | 0.055 ± 0.032 | 0.047 | 0.018 to 0.202 | 50.8 ± 15.6 | 47.6 | 24.9 to 97.2 |
| Zn | 2.73 ± 1.02 | 2.48 | 1.56 to 7.42 | see limits | 163 | 62.8 to 469 |
| Cu | 0.675 ± 0.228 | 0.660 | 0.232 to 1.18 | 47.8 ± 22.0 | 41.4 | 18.9 to 131 |
Soil values are captured as context only. The tuber chromium distribution is strongly skewed (median 0.023, mean 0.072, SD 0.153).
Methods and Evidence Fitness
At each of 56 fields, about 2 kg of soil (0 to 25 cm) was composited from seven points, and about 2 kg of tubers was taken at the matching points (PDF p.2 and p.3). Tubers were washed with distilled water, chopped, quartered and oven-dried at 65 °C for four days; peeling is not described. Soil (0.1 g) was digested with nitric and hydrofluoric acid; tuber material (0.5 g) with nitric and perchloric acid; all analyses were by ICP-MS in triplicate (PDF p.3). Accuracy was checked with soil reference materials GSS-14 and GSS-4 and with national reference materials carrot GSB-25 and tangerine GSB-11, but recoveries are not reported; parallel relative standard deviations were below 15%. The detection limit is given only for arsenic in tubers (0.01 mg/kg). Table 3 is reported on fresh weight, whereas the bioconcentration factors in Table 4 are on dry weight; the paper does not describe the conversion, but gives moisture per sample in a supplement (Table S2). The sampling year is not stated.
Evidence Fitness is Modeled or limited evidence. The tuber summary statistics have a stated basis, a reasonable sample size, ICP-MS with food reference materials and a stated arsenic detection limit, so they can stand as a regional occurrence description for field potatoes on high-background soils. They cannot be reconstructed into a distribution from the PDF, because individual results are only in the supplement (Table S5). Arsenic was not speciated, chromium was measured as total chromium, and cadmium and mercury were not measured. The data do not support any statement about potato starch or other processed potato products.
Assessment limits
The soil zinc mean is printed as 181 mg/kg in Table 2 but as 175 mg/kg in the Results and Discussion (PDF p.5 and p.8); the table value is recorded as rejected. The mean zinc bioconcentration factor is 0.087 in the abstract and Table 4 but 0.107 in the text (PDF p.1, p.6 and p.7), and Table 4 prints the zinc range as “0.010–226”, which is probably a typographical error for 0.226. In Table 6 (PDF p.8), the printed nickel daily intake divided by the stated reference dose gives about 5.9 × 10⁻³, not the printed hazard index of 5.92 × 10⁻²; the abstract’s upper hazard index carries the same discrepancy. The cluster table for soil chromium places 22 samples at or below 100 mg/kg, which implies 34 exceedances rather than the printed 35. Moisture in Table 1 is labelled as percent but printed as fractions (0.78 to 0.85). The potato “permissible maximum limits” in Table 3 are cited to a single reference without naming the standard. The text layer of PDF pages 3, 5 and 6 contains an overlaid draft; all numbers were taken from the rendered page images.
Provenance notes
The PDF (11 pages) was delivered in batch 05 of order HMI-GAP-INGEST-1 on 22 October 2026, filed in the manifest under the potato-starch folder. The SHA-256 matches the batch manifest. The PDF prints DOI 10.3390/foods11152379 and the title on PDF p.1; both match the manifest. The paper measures raw potato tubers, not potato starch, so it is routed to the potatoes ingredient page; it is not evidence for potato starch.
Related evidence
Update history
22 October 2026: Ingested from the batch-05 PDF. 45 tuber and soil records were added to the evidence register (21 tuber records, 23 soil context records, 1 rejected soil zinc mean).
Update history
No substantive edit history is available in this build. The full commit record is available in git.