Abstract

Urban soils generally have elevated metal contents originating from both point and diffuse pollution sources. Urban areas designated for children, who are most susceptible to any negative health effects of soil metals, may therefore have elevated soil metal contents. Children ingest soil both directly and by putting dirty hands and objects in their mouths. The soil ingested involuntarily mainly comprise very fine particles that have a larger surface area for sorption and may therefore hold higher metal contents than the bulk soil. However, only a fraction of the total soil metal content is generally bioaccessible, i.e. can be released from the soil and absorbed in the body. This thesis presents studies on total and bioaccessible metal contents in soils collected from playgrounds in urban Uppsala. Aqua regía digestion was used for total metal analysis while bioaccessibility was investigated for As, Cd, Cr, Ni and Pb using an in vitro digestion model simulating three ingestion scenarios: pica (compulsive ingestion of large amounts of soil), deliberate and involuntary ingestion. While mainly natural sources were found to contribute to soil contents of As, Al, Fe, Cr, Ni, Mn and W, the main sources of Cd, Cu, Hg, Pb and Zn were anthropogenic. Clay content and construction year of the playgrounds were more important factors for the soil metal content than past and present land use in the surrounding areas. The fine soil particles (<50 μm) had on average 150% higher metal contents than larger size fractions. Up to 250% enrichment was found in the fine fraction of sandy soils. The effect of particle size and soil mass on bioaccessibility was found to depend largely on soil texture, with the effects diminishing in soils with high clay content. Lead bioaccessibility was largely affected by gastrointestinal pH, which in turn was affected by particle size fraction and soil mass. The results emphasize the importance of soil analysis beyond only assessment of total metal content of bulk soil for a correct risk assessment of soil metals. The results also show the difficulties associated with generic soil metal guideline values in risk evaluation of soil contamination.