Small arms ranges are predominantly contaminated with lead from spent bullets. Models of risk assessment in these soils require a relative bioavailability term. The default value of 60% relative bioavailability can be replaced with a more meaningful site-specific value where supporting evidence from an in vivo swine-feeding study of the specific soils can be provided. The swine model has been used to successfully adjust the default bioavailability at some Superfund sites.The uniqueness of small arms ranges, in terms of their small size, potential number, and main pollutant being lead, makes it feasible to examine an alternative and more cost-effective estimate of potential exposure and toxicity. The in vitro surrogate of mammalian digestion provides a rapid, cost-effective, and robust alternative to the in vivo swine model. The objective of this project was to compare the bioavailability of soils from selected small arms ranges across the United States using the in vitro and in vivo assays and to validate the use of the rapid and low-cost in vitro technique for predicting oral bioavailability of lead at these sites. An additional objective was to assess the range of bioavailability in small arms range soils.

The relative bioavailability of lead in small arms range soils was compared using both the in vivo and in vitro models. Soil samples from eight small arms ranges were collected, dried, and sieved to a particle size of less than 250 micrometers. For the in vivo model, weanling pigs were orally dosed over a 15-day period and blood lead measurement was used to assess absorption from the gastrointestinal tract. Absorption was normalized to a lead acetate treatment group, giving the in vivo relative bioavailability. For the in vitro method, lead was extracted in a simulated gastric solution with a pH of 1.5 at 37º Celsius for 60 minutes, filtered, and analyzed for lead. This extractable lead was normalized to the total lead in the sample, giving the in vitro relative bioavailability.

A range of soil types from Maryland (2), Alaska, Louisiana, Nebraska, Washington, South Dakota, and Oregon were tested using both assays. All soils were sampled from berms and had high lead levels. For bioavailability measurements, the eight study soils had in vitro and in vivo means of 95±6% and 102±15%, respectively, showing excellent concurrence between the methods. In addition, x-ray fluorescence analysis showed a predominance of oxidized forms of lead, known to have high bioavailability, in these soils. The high bioavailability found in these samples prompted screening of soils from an additional 20 sites using the in vitro method alone, and these also had measured bioavailability greater than 90%. This project led to two major conclusions—the bioavailability of lead at all small arms ranges tested was greater than the default value of 60% (closer to 100%) and the in vivo and in vitro methods had a high degree of concurrence between them.

This project demonstrated the in vitro method to be a robust method capable of predicating bioavailability in the same range as the in vivo method, particularly when applied to small arms ranges. However, the expected range in bioavailability from a variety of small arms sites was not evident; therefore, assessing site-specific bioavailability may not be an important factor in risk assessment of ranges. These results indicate for the first time that small arms ranges have predominantly high lead bioavailability. (Project Completed - 2007)