SPATIAL PREDICTION OF ARSENIC BACKGROUND CONCENTRATIONS IN SOIL, OHIO, U.S.A

Natural and anthropogenic arsenic contained in soil is a potential health risk. Accordingly, various government entities are attempting to determine safe concentrations. Such guidelines are mainly based on toxicity modeling, but should also consider natural background concentrations. To address this issue, soil samples were collected to map natural arsenic (As) concentrations for the state of Ohio. Samples were collected at 348 locations (topsoil and subsoil), providing approximately one sample for every 290 km². Sites were selected to minimize the potential influence of anthropogenic inputs. Processing was conducted using whole sample, acid digestion and concentrations measured using atomic absorption. Arsenic concentrations ranged from 2 to 45.6 mg kg^-1. Geostatistical techniques were used to make a variety of maps. Risk assessment maps showing the probability of exceeding several concentration guidelines from the literature were made by indicator kriging. The mean concentration and variance at unsampled locations were estimated using sequential Gaussian simulation. Five areas of elevated (greater than the median value of 10 mg kg^-1) concentrations were identified and relationships to geologic parent materials, glacial sedimentation patterns, and soil conditions were tested statistically. Primarily, As spatial patterns match those of bedrock units known to contain Fe and As sulfide minerals, with secondary effects due to glacial sedimentation. Loss of As due to soil biogeochemical processes could not be ruled out for the lake plain of north west Ohio. The results show that As is a common element in Ohio and values exceeding 20 mg kg^-1 are not uncommon. Many concentration guidelines in the literature are well below natural background levels and therefore are unrealistic for application in Ohio as target levels for remediation activities.