Clair-Caliot et al. 2021 — Arsenic uptake by irrigated vegetables and cooked foods in Burkina Faso

This greenhouse study characterizes the transfer of arsenic from geogenic As-contaminated irrigation water to seven vegetables commonly grown and consumed in Burkina Faso (amaranth, carrot, green bean, lettuce, okra, spinach, tomato), and separately assesses how cooking with As-spiked or As-free water affects final As content in both greenhouse vegetables and locally purchased staple foods. Arsenic was measured as total As (tAs) by ICP-MS after HNO3/H2O2 digestion; speciation into inorganic vs organic arsenic was not performed. At the mid-point irrigation concentration of 500 µg/L As(V), edible-part concentrations were: spinach 6.6, lettuce 3.9, carrot 3.5, amaranth 2.2, okra 0.9, green bean 0.8, and tomato 0.2 µg/g dry mass. Cooking contaminated vegetables with As-free water reduced As content by an average of 39%, while cooking with As-contaminated water transferred As to the cooked food; steaming reduced As by 8- to 18-fold compared to boiling.

Key numbers

Speciation note: All As values in this study are total arsenic (tAs) measured by ICP-MS. No speciation into iAs and organic As was performed. Irrigation water As was applied as arsenate As(V) (sodium arsenate dibasic heptahydrate), so the arsenic entering plants was inorganic arsenate, but the measurement in plant tissue is total As and is recorded here as tAs per wiki convention.

Experimental design:

• 168 pots: 7 species × 4 As irrigation concentrations (0, 100, 500, 1,000 µg/L) × 6 replicates
• Greenhouse at 2iE, Ouagadougou; dry season November 2017–February 2018
• Culture period 65–130 days; harvested at maturity

As in edible parts at 500 µg/L irrigation water (µg/g DM):

• Spinach (leaves): 6.6 ± 0.5
• Lettuce (leaves): 3.9 ± 0.1
• Carrot (root): 3.5 ± <0.1
• Amaranth (leaves): 2.2 ± <0.1
• Okra (pods): 0.9 ± <0.1
• Green bean (pods): 0.8 ± <0.1
• Tomato (fruits): 0.2 ± <0.1

Average across all As-spiked irrigation levels (µg/g DM, edible parts):

• Spinach: 8.1 ± 5.6
• Lettuce: 3.6 ± 2.5
• Amaranth: 3.0 ± 2.5
• Carrot (all organs, edible): included in root-vegetable category
• Green bean: 1.5 ± 1.5
• Okra: 0.7 ± 0.4
• Tomato: 0.3 ± 0.2

Edible-part category averages (µg/g DM, across all spiked levels):

• Leafy vegetables (amaranth, lettuce, spinach): 4.9 ± 4.5
• Root vegetables (carrot): 2.9 ± 2.0
• Fruit/pod vegetables (tomato, okra, green bean): 0.8 ± 1.1

Cooking effects (greenhouse vegetables cooked in As-free water):

• Amaranth: 61% mean As extraction (up to 69%)
• Spinach: 3–31% As extracted (9% at 500 µg/L)
• Okra: 35–49% As extracted at 500–1,000 µg/L
• Overall average: 39% reduction from cooking with As-free water

Market food As content (uncooked, local Ouagadougou market):

• Local rice (Boulkiemdé): 0.127 µg/g tAs
• Maize flour (for tô): 0.06 µg/g tAs
• Black-eyed peas (bean/niébé): 0.04 µg/g tAs

Steaming vs boiling:

• Steamed yam: 7.9× lower As than boiled yam
• Steamed rice: 17.9× lower As than boiled rice (absorb-method)

Soil properties (experimental substrate):

• As: 0.684 µg/g, Cd: 0.029 µg/g, Pb: 6.620 µg/g
• Soil As was below EU maximum acceptable limit for agricultural soil (20 µg/g)

Translocation factors (TF = shoot As / root As):

• Only carrot had TF > 1 (leaves concentrate more than roots); all other species had TF < 1
• Tomato had lowest TF among measured species (~0.6)

Cancer risk (CR) for raw irrigated vegetables (Table 3, p. 10): CR exceeded 1 in 10,000 for green bean, lettuce, and spinach at 500 and 1,000 µg/L irrigation; for amaranth only at 1,000 µg/L; and for carrot at all spiked treatments (100, 500, and 1,000 µg/L). Tomato and okra had CR < 1 in 10,000 across all treatments. Risk characterization used EPA OSF for inorganic arsenic (1.5 (mg/kg/day)^-1) applied to total As — a conservative assumption.

Detection limits: 0.05 µg/L in digest liquid; 0.05 µg/g in food dry mass.

Methods (brief)

Greenhouse experiment at 2iE, Ouagadougou; sandy loam soil (subsurface 2–25 cm) spiked with As(V) as sodium arsenate dibasic heptahydrate. Plants harvested at maturity, freeze-dried, ground, digested in Burkina Faso by HNO3 at 80°C (Bhatti et al. 2013 protocol) or in Switzerland by HNO3/H2O2 at 240°C (ultraCLAVE microwave). Both methods compared for okra and spinach with good agreement for concentrations 0.5–20 µg/g. As determined by ICP-MS (Agilent 7500cx) at Eawag, Switzerland. Standard reference materials: NIST 1568b rice flour (118% mean recovery), NIST 1573a tomato leaves (125% mean recovery). Cooking experiments: greenhouse vegetables boiled in excess distilled water; market foods boiled, steamed, or prepared with cold water according to traditional Burkinabe recipes, in As-free or As-spiked cooking water. No arsenic speciation was performed; all values are total As.

Implications

Certification: This study is most relevant as evidence for the processing lever (cooking method) affecting final As in food, and for ranking vegetable types by relative As accumulation potential from contaminated irrigation water. The concentrations reported are from an experimental system with irrigated As concentrations far exceeding normal ambient conditions; they are not suitable as baseline contamination profile values for typical commercial production contexts. They establish ceiling values under high-As irrigation stress.

Courses: Excellent teaching case on irrigation-water-to-plant As transfer, translocation factors, differential accumulation by vegetable type (leafy > root > fruit/pod), and the effectiveness of cooking method (steaming vs boiling) as a mitigation strategy. The Burkina Faso context adds geographic equity dimension.

App: Not suitable for contamination_profile typical_ppb values — these are experimental high-dose exposure data, not field survey data from commercial vegetable markets. Use only for directional ranking (spinach and lettuce accumulate more As than tomato and okra under As-contaminated irrigation).

Microbiome: Not applicable.

Wiki pages updated on ingest

• leafy-vegetables
• spinach
• lettuce
• root-vegetables
• carrots
• tomato
• beans
• green-beans
• rice
• maize
• arsenic

Verification notes

2026-05-20 merge-enhance pass (v2 manual-fetch skill): preserved cite_key, license, access_url, and existing Key numbers (independently verified against PDF — concentrations at 500 µg/L, averages across spiked levels, category averages, cooking extraction percentages, steamed-vs-boiled ratios, soil properties, NIST recoveries, and detection limits all match source). Fixes applied: (a) raw_handle normalized from manual-fetch-kimi to MFK_uptake-of-arsenic-by-irrigated-vegetables-and-cook per current handle convention; (b) raw_path corrected — previous value was truncated mid-filename (“Cooked Food Product.pdf”) and now points to the actual file (“Cooked Food Products in Burkina Faso.pdf”); (c) ingredient slug tomatoes corrected to canonical tomato (taxonomy snapshot lists tomato as singular); (d) root-vegetables re-added to Wiki-pages-updated list (was in frontmatter but missing from body list); (e) removed dead wikilinks [[supply-chain/irrigation-water-arsenic]] and [[mitigation/cooking-arsenic-reduction]] — neither page exists in current taxonomy; (f) added near_duplicates: []; (g) updated advanced to 2026-05-20. Amaranth and okra (both extensively covered by this paper) are not yet in the ingredient taxonomy; their data is routed via the leafy-vegetables and fruit-vegetable umbrella slugs respectively per Part 5b broad-scope rule. Speciation discipline upheld: paper measured total As only, page records metals: [tAs] and Key numbers includes an explicit speciation note.

2026-05-20 audit-application (fresh-context subagent, verdict REVISE): two ⚠️ findings verified against PDF and applied. (1) Cancer-risk sentence corrected — prior wording said “leafy vegetables (amaranth, lettuce, spinach) and carrot under 500 and 1,000 µg/L” exceeded 1E-4. Verified against Table 3 (p. 10) and Discussion text (p. 8): amaranth CR at 500 µg/L is 9.5E-05 (below 1E-4), exceeds only at 1,000; green bean exceeded at 500 and 1,000 (was entirely missing from prior wording); carrot exceeded at all spiked treatments (100, 500, 1,000), not just 500 and 1,000. Sentence rewritten to match author wording exactly. (2) Added specific ingredient slugs green-beans (paper studies Phaseolus vulgaris pods explicitly in the greenhouse experiment) and maize (paper measures maize flour used for tô) — both exist in current taxonomy. Generic beans retained as umbrella for black-eyed pea/niébé market-food data. Carrots-vs-carrot slug choice noted by audit as inconsistent with tomato singular-form fix; not applied — both carrot and carrots exist in the taxonomy snapshot with comparable corpus usage (9 vs 12 sources), so the existing carrots declaration is preserved to avoid introducing slug churn.

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.