Weathering and dispersal of polychlorinated biphenyls from a known source in the Canadian Arctic

The results of a PCB weathering experiment (i.e., losses from soil and changes in PCB composition from soil over time) at Cambridge Bay, Northwest Territories, in the Canadian Arctic are presented. Nine plots representing three different general substrate types (dry barren, dry moss, and wet grass) were treated with three different PCB Aroclor formulations (1254, 1260, and a 1:1 mixture). Soil and plant samples were collected annually over a 5-year period and analyzed for individual PCB congeners. The results showed a high degree of variability between treatments that was ascribed to the presence of vegetation as well as non-homogenous PCB application and/or soil heterogeneity (e.g., biomass, organic carbon, and moisture content) within individual plots. Where readily interpretable data were available, the first-order rate constant for the loss of PCBs from the substrate was calculated to be approximately 0.5 yr(-1) (t(1/2) = 1.1 yr) (range = 0.3-1.0 yr(-1) for individual congeners). For the dry barren area, the rate of loss from soil of individual congeners was negatively correlated with the planar total surface area of the congener (or positively correlated with vapor pressure). For a wet grass area or a dry moss area, the rate of loss was not correlated with any of the standard physical constants, suggesting that live and detrital vegetation may play a key role in the release and retention of PCBs. The estimated loss rates for the three plots may or may not be applicable to PCB-contaminated soils in general, where PCB volatility may be reduced by the presence of hydrocarbon-based carriers or influenced by other factors not considered in this study.