Paper No. 9
Presentation Time: 4:00 PM
SOIL LEAD RISK ASSESSMENT IN URBAN GARDEN PLOTS
BUGDALSKI, Lauren, Department of Geology, Wayne State University, Detroit, MI 48202, LEMKE, Lawrence D., Department of Geology, Wayne State University, 0224 Old Main, 4841 Cass, Detroit, MI 48202 and MCELMURRY, Shawn P., Civil and Environmental Engineering, Wayne State University, Detroit, MI 48202, bc0254@wayne.edu
The rising popularity of urban gardening increases the importance of ensuring that food grown in urban environments with legacies of soil contamination from past anthropogenic activities is safe for human consumption. Soil sampling strategies for urban garden plots must be practical, cost effective, and adequate for risk assessment. To evaluate soil lead distributions and alternative sampling strategies for risk assessment, three former residential lots now functioning as urban garden plots in Detroit, Michigan, were investigated. Each lot was sampled using a nested soil sampling design to quantify meter-scale variability. Total and bioaccessible lead were measured using atomic absorption spectroscopy and modeled geostatistically. Monte Carlo analysis was then used to evaluate simulated sampling of mapped lead concentrations with the (potentially competing) objectives of: 1) characterizing the average soil lead concentration, or 2) detecting the presence of lead soil hotspots at each site. The use of individual and composited samples was examined.
Results of this analysis demonstrate that the goal of the soil lead sampling campaign affects risk assessment outcomes for reclaimed residential sites. In the first instance, if the objective is to determine a representative soil lead concentration for the entire site, increasing the number of individual samples taken improves the ability to estimate mean concentrations and the practice of compositing samples decreases the probability of a false positive (Type I error). Conversely, if the objective is to detect and possibly delineate hot-spots, increasing the number of individual samples increases the chance of hotspot detection, but compositing samples increases the probability of overlooking hotspots (Type II error). Current USEPA recommendations allow as few as two composite samples to estimate the average soil lead concentration on a standard sized residential lot. In sites where soil lead concentrations are nonuniform or hot spots are present, however, our results suggest that two composite samples may not be enough to characterize soil lead variability or fully assess the associated risks.
