Malone & Shakya 2024 — Trace metal contamination in US community garden soils

Malone and Shakya conducted an integrative literature review of trace metal contamination (Pb, As, Cd, Zn) in US urban community garden soils, synthesizing 52 peer-reviewed articles from Web of Science and PubMed. The central finding is that lead is the primary soil contaminant of concern in urban community gardens, driven by legacy industrial activity, leaded paint chips, and leaded gasoline residues. The US EPA’s residential screening level for lead — reduced from 400 ppm to 200 ppm in January 2024, and 100 ppm at residences with multiple lead sources — remains substantially higher than state, international, and recommended levels (California 80, WHO 85, Finland ~60, US EPA 2013 urban-gardener recommendation 100). Arsenic and cadmium are secondary concerns, often co-occurring with lead, with notable cadmium contributions from phosphate fertilizers, vehicle exhaust deposition, and treated wood. Zinc is generally low-toxicity to humans but functions as an indicator metal: where Zn is elevated, Pb/Cd/As are often also elevated. Environmental justice is a cross-cutting theme: contamination hotspots disproportionately affect lower-income and minority urban communities.

Key numbers

Soil concentrations by US city (Table 2, ppm; mean and range, or range only)

• Aspen, CO (former mine dump sites; 65 gardens, >195 samples): Pb 172 (9.2–808), As not reported, Cd 2.5 (0.2–14.2), Zn 120 (8.4–484)
• Baltimore, MD (urban; 104 gardens, 616 samples): Pb 104.5 (7.4–130.4), As 3.7 (0.2–13.5), Cd 1.4, Zn 139.7 (39.7–542)
• Boston, MA (traffic/industrial; 3 gardens, 88 plots): Pb 130 (117–170) in compost soil; As 30–39 in CCA-bordered garden plots; Cd, Zn not reported
• Cleveland, OH (vacant lots; 65 samples): Pb 224 (14–1241), As 15 (7–58), Cd 1.2 (0.5–2.5), Zn 197 (83–543)
• Detroit, MI (urban/residential; 2 gardens, 80 samples): Pb 151 (17–882); As, Cd, Zn not reported
• New Orleans, LA (urban/suburban backyard and community gardens; 27 gardens, approximately 600 samples): Pb 38.4 (1.4–9540), As 3 (0.7–61.7), Cd 0.318 (0.248–8.8), Zn 91.5 (17.8–7330)
• New York City, NY (urban/residential; 54 gardens, 564 samples): Pb 102 (11–2455), As 5.7 (<5.3–93.2), Cd <0.4 (<0.4–3.1), Zn 138 (21–2317)
• New York City, NY (urban/residential; 905 gardens, 1652 samples): Pb 600 (3–8912), As 12 (0.9–7.6), Cd 1.6 (0.1–11), Zn 327 (35–2352)
• Oakland, CA (urban/residential; 3 gardens, 6 samples): Pb 47–326 (range only); As, Cd, Zn not reported
• Philadelphia, PA (urban; 11 gardens, 78 samples): Pb 47.6–351.4, As 0.9–9.6, Cd 0.1–1.4, Zn 177.4–936 (ranges only)
• Philadelphia, PA (suburban; 5 gardens, 24 samples): Pb 10.3–185.5, As 0.77–3.22, Cd 0.2–0.5, Zn 39.2–158.5 (ranges only)
• Pittsburgh, PA (urban; 5 gardens, 20 samples): Pb 83.1–232.9, As 1.9–11.4, Cd 0.2–1.3, Zn 110.9–237.9 (ranges only)
• Roxbury and Dorchester, MA (backyard; 141 gardens, 692 samples): Pb 950 (80–3680); As, Cd, Zn not reported

Lead-in-soil regulatory thresholds (Table 1, all in soil, ppm)

• Finland (used in international context as approximation of European national means): 60
• California: 80
• World Health Organization: 85
• Canada: 140
• US EPA: 200 (residential, January 2024 update from 400; 100 ppm at residential properties with multiple lead sources)
• Connecticut: 400
• New York: 400

Additional non-Table-1 reference values cited in the paper: US EPA Technical Review Workgroups Lead Committee recommendation of 100 ppm for urban gardeners, never officially adopted by US EPA (the paper is internally inconsistent on year, naming 2013 on p. 5 and 2014 on p. 2); US EPA 400 ppm for children’s play areas (the pre-January-2024 residential standard); 1200 ppm for “all other areas of a yard” (likewise recently revised to 200 ppm); natural-background lead in uncontaminated US surface soils 22 ppm; California median-background arsenic level 5 ppm (the paper cites Clarke 2007 as ref [25] for this figure but does not supply an originating year for the background measurement).

Plant uptake and exposure findings cited from the reviewed studies

• Children absorb approximately 50% of ingested lead vs under 5% in adults; 72–91% of childhood lead exposure attributed to incidental soil ingestion; 20–80% of dust in urban homes is soil dust.
• Approximately 95% of plant lead is absorbed by roots; root vegetables (carrots, turnips, beets, radishes) are therefore more susceptible to lead accumulation than fruits or nightshade vegetables.
• Leafy vegetables (lettuce, cabbage) show elevated trace-metal accumulation in leaves, attributed to high transpiration.
• In one reviewed study, a majority of onions grown in untreated urban soil exceeded the EU lead standard for roots and crops, with 80–92% lead bioaccessibility; arsenic bioaccessibility in those soils was 93%, and approximately half of the sampled onions accumulated >0.1 mg/kg arsenic.
• In gardens bordered by chromated copper arsenate (CCA) treated wood, 68% of plots containing CCA-treated wood exceeded the California background arsenic level of 5 mg/kg; 30% of plots without CCA-treated wood also exceeded that background.
• Triple Super Phosphate fertilizer, manure compost, and raised-bed treatments increased arsenic extractability by 101%, 56%, and 114% respectively while reducing lead extractability — phosphate amendments and raised beds mobilize arsenic even as they immobilize lead.
• One reviewed study reported that compost decreased available soil lead by 8% but radishes grown in compost-amended soil had a 19% increase in lead content; spinach and dill grown in compost showed increased accumulation of Pb, Cu, Cd, and Cr.
• Lead in soils across cities with industrial legacies (Baltimore, Philadelphia, Pittsburgh, NYC) shows rural-to-urban gradients; Chicago metropolitan soil Pb is approximately 5× rural forest soil.
• Childhood blood lead: among children in urban areas, approximately 16% have BLLs above the CDC’s previous 10 µg/dL action level; the CDC reference value was lowered from 10 µg/dL to 5 µg/dL in 2012 and to 3.5 µg/dL in 2021.
• In Dorchester, MA, lead in urban raised-bed soils increased by 185 ppm over 4 years from wind-deposited contamination.

Environmental-justice findings cited from the reviewed studies

• Median household income from Baltimore urban-garden lead-cold-spots to lead-hot-spots declined from USD 75,692 to USD 38,757; hot-spot areas had a 74.9% increase in Black residents, a 14.2% increase in unemployment, and a 29.1% increase in SNAP enrollment.
• In Boston (24.2% Black or African American citywide), the neighborhoods with the highest childhood blood lead levels are 45% Black/African American in Dorchester and 53% Black/African American in Roxbury.
• Approximately USD 4000 to replace 15 cm of contaminated topsoil makes the most effective remediation economically inaccessible to low-income gardeners.

Methods (brief)

Integrative literature review. Web of Science and PubMed searched using the terms “community garden” and “urban garden”, refined with “contamination” or “metals” and “USA”. Initial pulls: 13,971 (community garden / WoS) and 3719 (community garden / PubMed); 8852 (urban garden / WoS) and 1824 (urban garden / PubMed). Sequential keyword refinement narrowed to 40 community-garden and 27 urban-garden articles, deduplicated to 52 for review. Two reviewers screened titles, abstracts, and full text. After excluding studies that did not measure trace metals in US urban community garden soils, 34 relevant primary studies remained; additional review-article references and government/website resources were included where they provided relevant context. No original measurements; this is a synthesis of previously-published soil contamination data. Time-window limitation: search performed in late 2023 / early 2024; the most-recent US EPA residential-soil screening update (January 2024, 400→200 ppm) is captured.

Limitations

This is a literature review, not original measurement work, so the underlying soil-concentration data inherit the heterogeneity of the 34 primary studies — different sampling depths, digestion methods, analytical instruments, and reporting conventions (mean only, mean with range, range only) are mixed in Table 2. Boston, Detroit, Oakland, and Roxbury/Dorchester rows in Table 2 report only Pb (no As/Cd/Zn). Several rows report range without mean (Oakland, both Philadelphia rows, Pittsburgh). Cadmium for Baltimore (1.4 ppm) is reported without a range. The Table 2 footnotes on the Boston entry clarify that the Pb value of 130 ppm refers to compost-amended soil and the As value of 30–39 ppm refers to garden plots bordered by CCA-treated wood, not whole-garden averages. The review explicitly excludes phytoremediation and does not synthesize crop-tissue concentrations against soil concentrations across the reviewed studies. No quantitative meta-analytic pooling is performed; the table is a descriptive summary. The review does not attempt percentile distributions or geographic-weighted exposure estimates.

Implications

For US urban garden soils, this review anchors the literature baseline that Pb is the dominant trace-metal contaminant of concern, with mean garden-soil Pb concentrations across the surveyed cities ranging from 38.4 ppm (New Orleans urban/suburban backyards) to 950 ppm (Roxbury and Dorchester, MA backyard gardens), and individual-sample maxima reaching 8912 ppm (NYC) and 9540 ppm (New Orleans). The 22 ppm natural-background-soil Pb gives an order-of-magnitude reference: mean urban-garden Pb in the reviewed cities exceeds background by roughly 2× (New Orleans) to 40× (Roxbury/Dorchester). Co-occurring As and Cd are reported in roughly half the city-level entries; Zn is consistently elevated where measured and functions as an indicator metal. The US EPA Technical Review Workgroups Lead Committee recommendation of 100 ppm for urban-gardener exposure scenarios — never adopted as the official screening level — is identified by the authors as the gap between the EPA’s current standard (200 ppm residential, post-January-2024) and the more protective levels used by California (80 ppm), the WHO (85 ppm), and European national systems (~60 ppm). For wiki purposes: this source contributes US soil-contamination context for the leafy-vegetables, root-vegetables, and onions pages — particularly for the soil-uptake-pathway sub-block — and for lead, arsenic, cadmium, and zinc urban-soil narratives. Phosphate fertilizer’s documented effect of mobilizing arsenic while immobilizing lead is a notable methodological finding for soil-remediation discussion.

Verification notes

• 2026-05-18 (Claude session, merge-enhance): Re-read full PDF (15 content pages plus references) against the prior page version (updated: 2026-05-14, raw_handle: papers-cube).
• raw_handle: corrected from the legacy placeholder papers-cube to the current convention PCMF_article-2-copy-6 (filename-derived).
• raw_sha256: added (8f5830a58035600fb7ec02f6840aa6ca1f31d37fa2ded752aaff9b6d93c05c3d).
• metals: corrected from [Pb, tAs, Cd] to [Pb, tAs, Cd, Zn]. Zn is one of the four analytes the paper reviews (named in abstract, Keywords, Table 2, and §3.4) and was incorrectly omitted in the prior page.
• matrices: corrected from [soil, vegetable, root-vegetable] to [soil]. The paper is a review of soil-contamination measurements from primary studies; it does not include original food-crop measurements. “vegetable” was not in the canonical matrices vocabulary; plant-uptake discussion is captured via ingredients slugs, which is the appropriate routing input.
• ingredients: changed from [carrots, leafy-vegetables, root-vegetables] to [leafy-vegetables, root-vegetables, onions]. carrots was dropped as the paper mentions carrots only as one example of a root vegetable susceptible to Pb uptake (single sentence on p. 9); root-vegetables is the appropriate broad-scope slug. onions added because the paper makes a specific numerical claim about onions (majority above EU Pb standard, 80–92% Pb bioaccessibility, half >0.1 mg/kg As accumulation).
• sample_population: corrected to reflect the paper’s actual two-database search strategy and the deduplication to 52 articles (the prior page said “52 peer-reviewed articles published through January 2024” but did not specify the search strategy or the WoS/PubMed split).
• Key numbers — Table 2 reconstructed in full. The prior page omitted Detroit, MI (151 ppm Pb mean; 2 gardens, 80 samples; urban/residential), the Philadelphia suburban row (Pb 10.3–185.5 ppm), the Pittsburgh row (Pb 83.1–232.9 ppm), and the Roxbury/Dorchester row (Pb 950 ppm mean, 80–3680 ppm range; 141 gardens, 692 samples). All four are now included.
• Key numbers — Boston, MA: prior page said “Pb 117–170 (garden a)“. Table 2 footnote a denotes “compost soil,” not “garden a.” Corrected to “Pb 130 (117–170) in compost soil” per Table 2 with footnote a; As 30–39 in CCA-bordered plots per footnote b.
• Key numbers — Aspen, CO: prior page omitted “As not reported”; Table 2 leaves As blank for Aspen. Now stated explicitly.
• Key numbers — Table 1 regulatory thresholds: added Connecticut (400) and New York (400) which the prior page omitted; clarified that the Finland row represents “a good approximation of the mean values of different national systems in Europe” per Table 1 footnote, not specifically Finland’s domestic standard.
• Key numbers — CCA-treated wood arsenic: prior page said “68% of plots exceeded 5 mg/kg As” without context. The paper (p. 7) actually reports two figures: 68% of plots WITH CCA-treated wood exceeded the 5 mg/kg California background, and 30% of plots WITHOUT CCA-treated wood also exceeded background. Both figures are now stated.
• Key numbers — onion finding: prior page said “Onions: >93% frequency of soil Cd contamination >0.6 mg/kg in reviewed sites; half of sampled onions showed >0.1 mg/kg Cd uptake.” This is wrong on the analyte (the paper discusses Pb and As for onions, not Cd) and conflates separate findings. Corrected per p. 9 of the paper: a majority of onions in the cited study exceeded the EU Pb standard for roots and crops after growth in untreated soil with 80–92% Pb bioaccessibility; arsenic bioaccessibility in those soils was 93%, and approximately half of sampled onions accumulated >0.1 mg/kg arsenic.
• Added the phosphate-fertilizer/raised-bed As-extractability finding (+101%, +56%, +114% respectively), the compost/radish lead-uptake finding (+19% in radish tissue despite −8% available soil lead), and the Dorchester raised-bed lead-increase finding (+185 ppm over 4 years from wind deposition) — all numerically specific findings the prior page summarized only narratively.
• Added the environmental-justice numbers from §6: Baltimore household-income gradient (USD 75,692 → 38,757 cold-spot to hot-spot), demographic deltas (74.9% Black resident increase, 14.2% unemployment increase, 29.1% SNAP increase), Boston neighborhood demographics for Dorchester (45%) and Roxbury (53%), and the USD 4000 topsoil-replacement cost. The prior page mentioned environmental justice as a “cross-cutting theme” without these specifics.
• Methods (brief) section added — the prior page conflated methodology with limitations under “Methods (brief)” but did not separately describe the search strategy. Now includes the database pulls (13,971/3719/8852/1824), the keyword-refinement sequence, the deduplication to 52, and the second-pass screen to 34 primary studies.
• Limitations section added — the prior page lacked this mandatory section. Now flags the heterogeneity of the underlying primary studies, the rows missing analytes in Table 2, the missing means/ranges for some rows, the Boston footnote interpretation, and the absence of meta-analytic pooling.
• Removed the legacy ”## Wiki pages updated on ingest” heading. Page routing is now handled entirely by the routing audit reading frontmatter; wiki-page-to-source-page references are not maintained in source-page body. Implications section retains the relevant page references inline ([[ingredients/...]], [[metals/...]]).
• Removed the prior ## Implications subsections “Certification:”, “Courses:”, and “App:” — Part 2 firewall guidance reserves Implications for what the paper contributes to the literature record, not what HMTc, course content, or the app should do with it. Implications now describes the source’s literature-baseline contribution without proposing certification thresholds or app modifiers.
• “Natural background US soil: 22 ppm Pb” retained, moved into the Plant-uptake/exposure-findings block where the paper actually cites it (p. 5).
• Audit subagent 2026-05-18 flagged two ⚠️ items in Check 1, both verified against the source and applied:
  • Subagent noted that the wiki page labeled the California background-arsenic value as “California 1991 background arsenic level 5 ppm” but the paper (p. 7) says only “California has elevated levels of arsenic in its soils … with a median background level of 5 ppm [25]” with no originating year. Verified: the “1991” attribution was a Claude error in the first ingest pass. Corrected to “California median-background arsenic level 5 ppm” with a note that the paper cites Clarke 2007 (ref [25]) without an originating year for the background measurement.
  • Subagent noted that the US EPA Technical Review Workgroups Lead Committee 100 ppm recommendation was dated “2013” in Key numbers and “2014” in Implications. Verified: the paper itself is internally inconsistent (p. 2 says “In 2014, they suggested…”, p. 5 says “In 2013, the US EPA Technical Review Workgroups Lead Committee created a recommendation of 100 ppm”). Reconciled to a single mention noting the paper’s own year inconsistency; removed the conflicting 2013/2014 phrasing from Implications.

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.