Background/Question/Methods
Lead contamination of the urban environment is often thought of as an antiquated problem, but legacies from historic lead use are very much a contemporary health concern.  In addition to the well-known health effects of lead-based paint contamination, lead dust and lead in soil can also pose human health risks.  Therefore, identifying where and how much lead persists in urban residential soil remains a goal of both ecologists and the public health community.  Ecologists have traditionally assumed that land use is a predictor of ecosystem services; however, this may not be the case for lead retention in urban soils.  Previous studies conducted as part of the Baltimore Ecosystem Study have shown that no relationship exists between lead levels in soil and land use.  In contrast, we hypothesize that land cover, which identifies individual landscape features, including soil, roads, buildings, and trees, better correlates with lead in soil because landscape features have varying abilities to sequester lead.  To test this hypothesis, soil data for 125 random plots located in Baltimore were spatially joined to a new urban classification, HERCULES (High Ecological Resolution Classification for Urban Landscapes and Environmental Systems), that focuses on land cover.
Results/Conclusions

Within a 30m radius buffer individual landscape features, as defined by HERCULES, were delineated to examine correlations between landscape features and lead levels in soil.  Regression analyses showed a significant positive relationship between lead concentrations in soil and building (p = 0.0287) and pavement (p = 0.0078) densities.  Bare soil, coarse vegetation, and fine vegetation showed no relationship.  These analyses indicate that some individual features of land cover are better predictors of lead retention in urban soils than land use.  To test this hypothesis, detailed soil sampling stratified by landscape features that are predicted to affect lead retention is underway.  To date, 14 residential properties in Baltimore City have been sampled using a USEPA approved x-ray fluorescence (XRF) multi-element spectrum analyzer which allows for efficient in-field soil sampling for lead concentration.  Of the properties sampled, 21% had average XRF lead values that exceed the USEPA reportable limit of 400 ppm and 50% had at least one reading that exceeded the USEPA reportable limit.  Future work will focus on the development of a spatially explicit model used to predict levels of lead in urban residential soil.