Assessing the influence of confounding biological factors when estimating bioaccumulation of PCBs with passive samplers in aquatic ecosystems

摘要

Passive samplers are promising surrogates for organisms, mimicking bioaccumulation. However, several biological characteristics disturb the passive partitioning process in organisms by accelerating or restraining bioaccumulation, resulting in species-specific body residues of hydrophobic organic contaminants (HOCs). In addition to site-specific characteristics and HOC concentrations, age, sex, diet, biotransformation capability and habitat-specific characteristics may affect body residues. Two passive sampler types, polyethylene (PE) and polydimethylsiloxane (PDMS) were deployed in a PCB-con-taminated freshwater lake water and sediment, respectively, to assess their bioaccumulation prediction capacity. In order to understand the importance of biological characteristics in the bioaccumulation process, we explored bioaccumulation in biota from plants and plankton to mussels and fish. The PCB concentrations in the PE sheet reflected the bioavailable concentration of PCBs slightly better than those in the PDMS samplers. Passive samplers were good predictors of PCB concentrations in fish, whereas concentrations in algae and invertebrates were overestimated. When comparing the measured concentrations in biota to the estimated concentrations using the PE samplers, the average regression slope was 0.87 for all biota and 1.22 for fish, and average modeling efficiency (EF) was 3.02 for all biota and 0.6 for fish. The best model performance was achieved for fish in trophic levels 3-4. Bioaccumulation was species-specific and dependent on the trophic level and diet Closer examination revealed that metabolic capability changes during the life span, and source of nutrition determined the biomagnification of HOCs, which differed between the fish species. Thus, species composition and available prey selection compose a unique bioaccumulation scenario and the resulting body residues. Due to the existing variation in body residues derived from passive samplers, extrapolating the results from one to another ecosystem must be done with caution. Passive samplers, however, offer a very powerful tool for risk assessment on the ecosystem level.