Testing urban soils for lead: Community-based use of the incremental sampling method to increase data quality and accessibility

Abstract:

Elevated lead (Pb) concentrations in urban soils create a hazard for communities engaging in urban agriculture. Mitigating risk to these communities is challenging because soil testing is expensive, translating results requires specialized knowledge, and the process is often considered inaccessible. Furthemore, conventional soil sampling methods (combining 1-5 subsamples into a composite sample) used in urban agriculture might not accurately represent true mean lead concentrations, as soil lead distributions in urban soils can be highly heterogeneous. Measuring, communicating, and mitigating lead exposure risk remains a challenge at the community level.

Working with low-income communities of color in the San Francisco East Bay (California, USA), we conducted in-person workshops to train community members to sample soils using an incremental sampling method (ISM). Samples collected with ISM can represent means of heterogeneous distributions more accurately than can discrete samples or composite samples combining few subsamples. Participants collected triplicate incremental samples (each combining 30 subsamples into one sample) from each of 32 sites. After testing, participants received soil reports and participated in workshops on interpreting results. Researchers shared recommendations for best practices based on lead concentrations.

Lead concentrations in 17 sites were above the California EPA residential soil screening level (80 mg/kg). Across sites, lead concentrations (study mean±standard error 178.2±34.2 mg/kg) reflected expected lead variability in urban soils. Relative standard error in triplicate samples was below 10% in 18 sites and 20% in 28 sites, suggesting that these data were precise and reproducible, or that our modified ISM biased results. Though less than half of workshop participants completed sampling, those who did reported that their high degree of confidence in results justified the effort expended to collect samples. Based on the workshops, we developed a framework to communicate participatory risk-based decision making in the absence of standard thresholds coupling soil lead concentrations to health outcomes. Overall, we found ISM is suitable for community use but requires careful communication of soil lead variability and exposure risk.