On the reversibility of environmental contamination with persistent organic pollutants.

An understanding of the factors that control the time trends of persistent organic pollutants (POPs) in the environment is required to evaluate the effectiveness of emission reductions and to predict future exposure. Using a regional contaminant fate model, CoZMo-POP 2, and a generic bell-shaped emission profile, we simulated time trends of hypothetical chemicals with a range of POP-like partitioning and degradation properties in different compartments of a generic warm temperate environment, with the objective of identifying the processes that may prevent the reversibility of environmental contamination with POPs after the end of primary emissions. Evaporation from soil and water can prevent complete reversibility of POP contamination of the atmosphere after the end of emissions. However, under the selected conditions, only for organic chemicals within a narrow range of volatility, that is, a logarithm of the octanol air equilibrium partition coefficient between 7 and 8, and with atmospheric degradation half-lives in excess of a few month can evaporation from environmental reservoirs sustain atmospheric levels that are within an order of magnitude of those resulting from primary emissions. HCB and α-HCH fulfill these criteria, which may explain, why their atmospheric concentrations have remained relatively high decades after their main primary emissions have been largely eliminated. Soil-to-water transfer is found responsible for the lack of reversibility of POP contamination of the aqueous environment after the end of emissions, whereas reversal of water-sediment exchange, although possible, is unlikely to contribute significantly. Differences in the reversibility of contamination in air and water suggests the possibility of changes in the relative importance of various exposure pathways after the end of primary emissions, namely an increase in the importance of the aquatic food chain relative to the agricultural one, especially if the former has a benthic component. Since simulated time trends were strongly dependent on degradation half-lives, partitioning properties and selected environmental input parameters, it should not be surprising, that different field studies often generate highly divergent time trends.